A method and device for joint optimization of drone position and resource allocation

By detecting user location and cluster classification method to determine the drone location and optimize resource allocation using the successive convex approximation method, the problem of unreasonable drone location and resource allocation in the existing technology is solved, and efficient edge computing services are realized and high-precision sensing.

CN118250821BActive Publication Date: 2024-11-12NANJING UNIV OF POSTS & TELECOMM
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410350252.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-03-26
Publication Date
2024-11-12
Estimated Expiration
2044-03-26

AI Technical Summary

Technical Problem

The existing technology cannot effectively combine drone location and resource allocation, resulting in the inability to meet the needs of delay-sensitive applications. The sensing and communication separation design cannot achieve high data transmission rate and high-precision sensing at the same time, and cannot meet the new intelligent applications. Require.

Method used

By detecting the user's location, determining the drone's location based on the Kmeans++ cluster classification method, and using the successive convex approximation method to optimize the user's transmission power and local calculation frequency, repeatedly calculate the user data upload time until the change in the user equipment cost function value is less than the threshold, realizing Joint optimization of resource allocation and drone location.

Benefits of technology

It reduces the cost of user equipment completing tasks, improves system performance, meets the needs of edge computing services, and realizes high data transmission rate and high precision sensing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118250821B_ABST
    Figure CN118250821B_ABST
Patent Text Reader

Abstract

The present invention discloses a joint optimization method and device for the position and resource allocation of unmanned aerial vehicles. Under the constraints of the amount of task data of user equipment, communication and calculation causality, a user cost function minimization problem is established, which optimizes the user equipment transmission power, local calculation frequency, data upload time and the position of the unmanned aerial vehicle. The proposed method first uses the Kmeans++ algorithm to optimize the position of the unmanned aerial vehicle, and then decomposes the resource allocation problem into the user transmission power and local calculation frequency optimization sub-problems and the data upload time optimization sub-problems. The user transmission power and local calculation frequency sub-problems are solved by a successive convex approximation algorithm, and the data upload time expression is obtained by deduction, and finally the appropriate user transmission power, local calculation frequency, data upload time and unmanned aerial vehicle position are obtained. The present invention reduces the cost of the user equipment to complete the task by reasonably arranging resources and the position of the unmanned aerial vehicle flight.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to a method and device for joint optimization of unmanned aerial vehicle position and resource allocation, belonging to the technical field of communications. Background Art

[0002] In recent years, computing-intensive or latency-sensitive applications (such as video processing, augmented reality, etc.) have developed rapidly, which has put forward new requirements for the computing power of terminals. Current user terminals cannot meet the needs of various services, and generally offload them to computing centers for completion. However, computing centers are generally far away from users and cannot meet latency-sensitive applications. Therefore, mobile edge computing technology has emerged. Mobile edge computing (MEC) deploys a large number of edge devices at the edge of the mobile communication network to provide idle computing power and storage resources to help user devices perform tasks. There are many types of edge devices, such as ground base stations, vehicle-mounted base stations, and drones. The location of ground edge devices is usually fixed and cannot be changed or moved according to user needs, which limits the capabilities of edge devices. Due to the mobility, simple configuration, low cost, and small size of drones, it can provide users with better edge computing services as an edge computing node. The drone-assisted MEC system provides services to users in an energy-saving and low-latency manner, improving the performance of the system. With the development of mobile communication technology, new intelligent applications and services such as machine-type communication and unmanned driving have emerged. The separation of sensing and communication cannot achieve high data transmission rate and high-precision sensing at the same time, nor can it meet the requirements of these new services. Summary of the invention

[0003] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a method and device for joint optimization of drone position and resource allocation. The method is aimed at the scenario where the drone acts as an edge node to provide edge computing services to users within the coverage area, and the user uses the integrated wireless signal to partially unload the task to the drone and uses the integrated signal echo to perceive the surrounding environment. By reasonably arranging resources and the flight position of the drone, the cost of the user device to complete the task is reduced.

[0004] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0005] In a first aspect, the present invention provides a method for jointly optimizing the position and resource allocation of a drone, comprising:

[0006] S1. Detect user location;

[0007] S2, clustering the users based on the user locations and the preset number of drone hovering time slots, and determining the drone locations based on the clustering results;

[0008] S3. Initialize the user data upload time according to the drone position and obtain the initial value of the upload time;

[0009] S4, determining the user transmission power and local calculation frequency based on the drone position and the initial value of the upload time;

[0010] S5. Recalculate the user data upload time according to the drone position, user transmission power and local calculation frequency;

[0011] S6, repeating steps S4 and S5 until the change in the cost function value of the user equipment is less than a specified threshold;

[0012] S7. Allocate resources to the user according to the user transmission power, local computing frequency and user data upload time obtained in steps S3-S6, so that the user uses the allocated resources to offload tasks, and the drone moves at the drone position obtained in step S2 and provides computing services to the user.

[0013] Furthermore, in step S2, clustering the users based on the user positions and the preset number of drone hovering time slots, and determining the drone positions according to the clustering results, including:

[0014] S2.1, randomly select a user location as the first initial cluster center;

[0015] S2.2. Calculate the distance of all users from the currently determined nearest cluster center according to the formula Calculate the probability of all users’ positions becoming the next initial cluster center, where m represents the sequence number of user m;

[0016] S2.3. Determine the new cluster center by the roulette method. Divide the interval [0,1] into q intervals according to the probability of each user position becoming the next initial cluster center, where q is the total number of users. Then randomly generate a random number between 0 and 1. The user position corresponding to the interval where the random number is located is the new cluster center position.

[0017] S2.4, repeat steps S2.2-S2.3 until the remaining N-1 initial cluster centers are selected, where N is the number of drone hovering time slots;

[0018] S2.5, grouping users whose nearest cluster center is the same cluster center into the same cluster;

[0019] S2.6. The average value of the horizontal coordinates of all user positions in each cluster is used as the horizontal coordinate of the new cluster center, and the average value of the vertical coordinates of all user positions is used as the vertical coordinate of the new cluster center;

[0020] S2.7, repeat steps S2.5-S2.6 until the change of all cluster centers is less than the threshold or the maximum number of cycles is reached;

[0021] S2.8. The hovering height of the drone is obtained as the maximum value of the lowest height covering all user devices in the current cluster and the lowest height of the drone;

[0022] S2.9, the first position where the drone hovers is the position of the cluster center closest to the coordinate origin, and the cluster center position closest to the current drone position in the untraversed cluster center set is the next hovering position;

[0023] S2.10. Get the final position of the UAV in each hovering time slot.

[0024] Furthermore, in step S3, the user data upload time is initialized according to the drone position to obtain the initial value of the upload time, including:

[0025] The hovering time slot number of the drone when user m communicates with the drone is S m , initialize the user data upload time to half of the hovering time slot length, and obtain the initial value of the upload time;

[0026] The hovering time slot length is calculated using the following formula:

[0027]

[0028] in, is the length of the nth hovering time slot, is the total hovering time slot length;

[0029] in, is the length of the nth flight slot, and the formula is: Among them, v is the speed of the drone when flying, L UAV (n) is the position of the UAV in hovering time slot n.

[0030] Furthermore, in step S4, determining the user transmission power and the local calculation frequency based on the drone position and the initial value of the upload time includes:

[0031] S4.1. Calculate the distance between the user device m and the drone in the hovering time slot n: Among them, L m is the position of user equipment m during the entire flight period;

[0032] The channel gain between user equipment n and the drone in hovering time slot n is obtained as:

[0033]

[0034] Where ξ0 is the reference channel gain when the distance between the user and the drone is 1, and θ is the path loss factor;

[0035] Calculate the total number of user devices communicating with the drone during the hovering time slot n as The available bandwidth of the user equipment communicating with the UAV in the hovering time slot n is The information transmission rate of user equipment m in the nth hovering time slot is obtained as:

[0036]

[0037] Where N0 is the power spectral density of the channel additive white Gaussian noise, P m (n) is the transmit power of user equipment m in hovering time slot n;

[0038] S4.2. When the drone is in a hovering state, the user equipment within the drone's communication range performs local computing and data upload at the same time. The drone's computing power is F. The drone distributes the drone's computing resources evenly to all users within the coverage area during the hovering time slot. When the drone is in a flying state, or the user equipment is not within the drone's communication range, the user equipment can only perform local computing. The local computing frequency of user equipment m in the nth hovering time slot is The local calculation frequency in the nth flight slot is Therefore, the amount of data calculated by user equipment m in the nth hovering time slot is Among them, D m represents the computational intensity of the task of user equipment m; the amount of data calculated in the nth flight time slot is The local computing energy consumption of user equipment m in the nth hovering time slot is The local computation energy consumption in the nth flight time slot is where β loc Effective capacitance coefficient;

[0039] S4.3. Calculate the conditional mutual information of user equipment m in hovering time slot n as:

[0040]

[0041] Among them, s m (n) represents the waveform used by user equipment m to transmit data in hovering time slot n; φ m (n) represents the impulse response of the channel between user equipment m and the sensing target in hovering time slot n; z m (n) represents the reflected echo signal received by user equipment m in hovering time slot n; T s Indicates the duration of an OFDM symbol, Q m(n) represents the Fourier transform of the impulse response of the channel between the user device m and the sensing target in the hovering time slot n;

[0042] S4.4. Calculate the total energy consumption of all user devices in all time slots as:

[0043]

[0044] in, is the energy consumed by user equipment m in the nth hovering time slot, Indicates the length of time that user equipment m transmits data;

[0045] The total conditional mutual information of the user equipment in all time slots is:

[0046]

[0047] S4.5. Initialize the stopping condition of the successive convex approximation method, Taylor expansion point

[0048] S4.6. Based on the determined initial values ​​of the drone position and upload time, establish the user cost minimization problem P1, which is specifically expressed as follows:

[0049] P1:

[0050] st(C1)

[0051] (C2)

[0052] (C3)

[0053] (C4)

[0054] (C5)

[0055] in, is the user device set, expressed as is the hover time slot index set, expressed as The flight slot index set is represented as

[0056]

[0057] S4.7. Perform a first-order Taylor expansion on the left side of the constraint (C2) inequality of problem P1 to obtain the optimization problem P2, which is specifically expressed as:

[0058] P2:

[0059] st(C1)

[0060] (C2)

[0061] (C3)

[0062] (C4)

[0063] (C5) in

[0064]

[0065]

[0066] S4.8. Use the CVX toolbox to solve problem P2 and update the Taylor expansion point to the solution of problem P2;

[0067] S4.9. Repeat steps S4.7-S4.8 until the objective function value of problem P2 meets the stopping condition of the successive convex approximation method.

[0068] S4.10. After the loop is completed, the solution to problem P2 is used as the user transmission power and the local calculation frequency.

[0069] Furthermore, in step S5, the user data upload time is recalculated according to the drone position, the user transmission power and the local calculation frequency, including:

[0070] S5.1. Establish the optimization problem P3 based on the determined drone position, user transmission power and local computing frequency, which is specifically expressed as follows:

[0071] P3:

[0072] st(C1)

[0073] (C2)

[0074] (C3)

[0075] (C4)

[0076] (C5)

[0077] When n=S m season f Taking partial derivatives we get:

[0078]

[0079] when When f Monotonically increasing, the optimal value is when When Monotonically decreasing, the optimal value is

[0080] S5.2. The solution to problem P3 is used as the user data upload time.

[0081] Furthermore, in step S6, the formula for calculating the cost function value of the user equipment is: cost = E total -υMI sum .

[0082] In a second aspect, the present invention provides a device for jointly optimizing the position and resource allocation of a drone, comprising:

[0083] A detection module, used to detect the user's location;

[0084] A clustering module is used to cluster users based on user locations and a preset number of drone hovering time slots, and determine the drone location based on the clustering results;

[0085] A first calculation module is used to initialize the user data upload time according to the drone position, obtain the initial value of the upload time, and determine the user transmission power and the local calculation frequency based on the drone position and the initial value of the upload time;

[0086] The second calculation module is used to recalculate the user data upload time according to the drone position, the user transmission power and the local calculation frequency;

[0087] An iteration module, configured to repeatedly execute the first calculation module and the second calculation module until the change in the cost function value of the user equipment is less than a specified threshold;

[0088] The resource allocation module is used to allocate resources to users according to the obtained user transmission power, local computing frequency and user data upload time, so that users use the allocated resources to unload tasks, and the drone moves on the drone position obtained in the clustering module and provides computing services for users.

[0089] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the aforementioned methods.

[0090] In a fourth aspect, the present invention provides a computer device / equipment / system, comprising:

[0091] Memory, for storing computer programs / instructions;

[0092] A processor is used to execute the computer program / instructions to implement the steps of any of the aforementioned methods.

[0093] In a fifth aspect, the present invention provides a computer program product, comprising a computer program / instruction, which implements the steps of any of the aforementioned methods when executed by a processor.

[0094] Compared with the prior art, the present invention has the following beneficial effects:

[0095] 1. The present invention considers drones as edge computing nodes to provide edge computing services to users within the coverage area. Users use the integrated wireless signal to offload computing tasks to drones in a partial offloading manner, and use the integrated signal echo to perceive the surrounding environment. Considering the requirements of user data volume, perception performance, and drone location, the weighted mutual information of user energy consumption is used as the user's cost.

[0096] 2. The present invention first determines the position of the UAV by using the Kmeans++ clustering classification method, and then decomposes the optimization problem into two sub-problems when the position of the UAV is determined and solves them by an alternating iterative method.

[0097] 3. The present invention adopts the successive convex approximation method to obtain the user transmission power and the local calculation frequency, and obtains the relationship expression between the data upload time and the transmission power and the local calculation frequency through mathematical derivation. After the optimal values ​​of the user transmission power and the local calculation frequency are determined, the user data upload time can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0098] Figure 1 is a flow chart of a method for joint optimization of drone location and resource allocation provided by an embodiment of the present invention;

[0099] Figure 2 is a schematic diagram of a system model provided by an embodiment of the present invention; DETAILED DESCRIPTION

[0100] 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.

[0101] Example 1

[0102] This embodiment introduces a joint optimization method for drone position and resource allocation, including:

[0103] S1. Detect user location;

[0104] S2, clustering the users based on the user locations and the preset number of drone hovering time slots, and determining the drone locations based on the clustering results;

[0105] S3. Initialize the user data upload time according to the drone position and obtain the initial value of the upload time;

[0106] S4, determining the user transmission power and local calculation frequency based on the drone position and the initial value of the upload time;

[0107] S5. Recalculate the user data upload time according to the drone position, user transmission power and local calculation frequency;

[0108] S6, repeating steps S4 and S5 until the change in the cost function value of the user equipment is less than a specified threshold;

[0109] S7. Allocate resources to the user according to the user transmission power, local computing frequency and user data upload time obtained in steps S3-S6, so that the user uses the allocated resources to offload tasks, and the drone moves at the drone position obtained in step S2 and provides computing services to the user.

[0110] Reference Figure 1 , Figure 2 The application process of the joint optimization method for drone location and resource allocation provided in this embodiment specifically involves the following steps:

[0111] S1, obtain the drone computing resources, bandwidth resources, user location, channel status, and mission data volume;

[0112] S2, clustering users using Kmeans++ algorithm based on user locations and preset drone hovering time slots, and determining drone locations based on clustering results;

[0113] S2.1, randomly select a user location as the first initial cluster center;

[0114] S2.2. Calculate the distance of all users from the currently determined nearest cluster center, and then use the formula Calculate the probability of all users’ positions becoming the next initial cluster center, where m represents the sequence number of user m;

[0115] S2.3. Determine the new cluster center by the roulette method, that is, divide the interval [0,1] into q intervals according to the probability of each user position becoming the next initial cluster center, where q is the total number of users, and then randomly generate a random number between 0 and 1. The user position corresponding to the interval where the random number is located is the new cluster center position;

[0116] S2.4, repeat steps S2.2-S2.3 until the remaining N-1 initial cluster centers are selected, where N is the number of drone hovering time slots;

[0117] S2.5, grouping users whose nearest cluster center is the same cluster center into the same cluster;

[0118] S2.6. The average value of the horizontal coordinates of all user positions in each cluster is used as the horizontal coordinate of the new cluster center, and the average value of the vertical coordinates of all user positions is used as the vertical coordinate of the new cluster center;

[0119] S2.7, repeat steps S2.5-S2.6 until the change of all cluster centers is less than the threshold or the maximum number of cycles is reached;

[0120] S2.8. The hovering height of the drone is the maximum value of the lowest height covering all user devices in the current cluster and the lowest height of the drone;

[0121] S2.9. The first position where the drone hovers is the position of the cluster center closest to the coordinate origin. The cluster center position closest to the current drone position in the untraversed cluster center set is the next hovering position.

[0122] S2.10, obtaining the final position of the UAV in each hovering time slot;

[0123] S3, initialize the user data upload time according to the drone position;

[0124] S3.1. The flight cycle of the drone is T. The entire flight cycle is divided into N hovering time slots and N-1 flight time slots. The hovering time slot index set is expressed as The flight slot index set is represented as The length of the nth hovering time slot is expressed as The length of the nth flight slot is expressed as The position of user equipment m during the entire flight period is represented by L m =[x m ,y m ,0], the position of the drone in the hovering time slot n is represented by L UAV (n) = [x UAV (n),y UAV (n),H UAV (n)], the UAV flies at a constant speed v in a straight line, and the length of the nth flight time slot is The total hovering time slot length is The length of the nth hovering time slot is proportional to the number of users communicating with the drone in the current time slot, expressed as

[0125] S3.2, the drone hovering time slot number of user m communicating with the drone is Sm , the initialization user data upload time is half of the hovering time slot length, that is,

[0126] S4, determining the user transmission power and local calculation frequency using a successive convex approximation method based on the initial value of the drone position and upload time;

[0127] S4.1. The user equipment set is represented as Each user device has a computing task to be processed, which can be divided into two parts. One part is calculated locally by the user device using its own processor, and the other part can be transmitted to the drone through wireless communication. The drone uses its own processor to assist the user in processing. The user device can communicate with the drone when the drone is in a hovering state and is within the drone's communication coverage. The user device uses the communication perception integrated signal when communicating with the drone, and uses the echo of the integrated signal to perceive the surrounding environment while transmitting data.

[0128] S4.2. The task to be processed by the user device uses a tuple <I m ,D m > indicates that I m Indicates the data size of the task of user device m, in bits, D m It indicates the computational intensity of task m of user device, that is, the number of processor cycles required to calculate 1 bit of task data, in cycles / bit.

[0129] S4.3 The communication coverage of the drone is in the shape of a cone. The section of the cone passing through the vertex and perpendicular to the ground is an isosceles triangle. The angle between the hypotenuse of this triangle and the altitude perpendicular to the base is denoted as γ. In the hovering time slot n, the user equipment within the coverage of the drone communicates using FDMA. The total bandwidth that can be allocated to the drone is B, which is evenly distributed to all user equipment that need to unload data within the coverage area. The transmission power of user equipment m in the hovering time slot n is P m (n), the distance between the user device m and the drone in the hovering time slot n is The wireless communication channel model considers the line-of-sight wireless channel model. The channel gain between user device n and the drone in hovering time slot n is

[0130]

[0131] Where ξ0 is the reference channel gain when the distance between the user and the drone is 1, and θ is the path loss factor.

[0132] The total number of user devices communicating with the drone during the hovering time slot n is The available bandwidth of the user equipment communicating with the UAV in the hovering time slot n is The information transmission rate of user equipment m in the nth hovering time slot is:

[0133]

[0134] Where N0 is the power spectral density of the channel additive white Gaussian noise.

[0135] In the hovering time slot n, the length of time that user equipment m transmits data is The length of time that the drone helps the user device m to calculate the task is The amount of data uploaded by user equipment m in the nth hovering time slot is The energy consumed is

[0136] S4.4. When the drone is in a hovering state, the user equipment within the drone's communication range performs local computing and data upload at the same time. The drone's computing power is F, and the drone distributes the drone's computing resources evenly to all users within the coverage area during the hovering time slot. When the drone is in a flying state, or the user equipment is not within the drone's communication range, the user equipment can only perform local computing. The local computing frequency of user equipment m in the nth hovering time slot is The local calculation frequency in the nth flight slot is Therefore, the amount of data calculated by user equipment m in the nth hovering time slot is The amount of data calculated in the nth flight slot is The local computing energy consumption of user equipment m in the nth hovering time slot is The local computation energy consumption in the nth flight time slot is where β loc The effective capacitance coefficient depends on the structure of the processor chip.

[0137] S4.5. Use m (n) represents the waveform used by user equipment m to transmit data in hovering time slot n. m (n) represents the impulse response of the channel between user equipment m and the sensing target in hovering time slot n. m (n) represents the reflected echo signal received by user equipment m in hovering time slot n. The conditional mutual information of user equipment m in hovering time slot n is

[0138]

[0139] Where T s Indicates the duration of an OFDM symbol, Q m(n) represents the Fourier transform of the impulse response of the channel between the user equipment m and the perceived target in the hovering time slot n, and its value is related to the antenna gain, radar cross section and channel gain of the transmission and reception process.

[0140] S4.6, the energy consumption of user equipment includes the energy consumption of uploading data, the local computing energy consumption when the drone is in the hovering state, and the local computing energy consumption when the drone is in the flying state. The total energy consumption of all user equipment in all time slots is

[0141]

[0142] The total conditional mutual information of the user equipment in all time slots is

[0143]

[0144] S4.7. Initialize the stopping condition of the successive convex approximation method, Taylor expansion point

[0145] S4.8. Based on the determined initial values ​​of the drone position and the user upload time, the user cost minimization problem P1 is established, which is specifically expressed as follows:

[0146] P1:

[0147] st(C1)

[0148] (C2)

[0149] (C3)

[0150] (C4)

[0151] (C5)

[0152] S4.9. Perform a first-order Taylor expansion on the left side of the constraint (C2) inequality of problem P1 to obtain the optimization problem P2, which is specifically expressed as:

[0153] P2:

[0154] st(C1)

[0155] (C2)

[0156] (C3)

[0157] (C4)

[0158] (C5) in

[0159]

[0160]

[0161] S4.10. Use the CVX toolbox to solve problem P2 and update the Taylor expansion point to the solution of problem P2.

[0162] S4.11, repeat steps S4.9-S4.10 until the objective function value of problem P2 satisfies the stopping condition of the successive convex approximation method;

[0163] S4.12. After the loop is completed, the solution to problem P2 is used as the user transmission power and the local calculation frequency.

[0164] S5. Recalculate the user data upload time according to the drone position, user transmission power and local calculation frequency;

[0165] S5.1. Establish the optimization problem P3 based on the determined drone position, user transmission power and local computing frequency, which is specifically expressed as follows:

[0166] P3:

[0167] st(C1)

[0168] (C2)

[0169] (C3)

[0170] (C4)

[0171] (C5)

[0172] The optimization problem P3 is a convex optimization problem. This paper uses mathematics to directly derive the optimal solution of problem P3. Proof: According to constraint (C5), when n≠S m hour, When n=S m season f Taking partial derivatives we get

[0173]

[0174] It can be seen that the function f is a monotonic function, and by transforming the constraint (C1), we can get

[0175]

[0176] By constraining (C2) to transform, we can get

[0177]

[0178] By constraining (C3) to deform, we can get

[0179]

[0180] therefore

[0181]

[0182] when When f Monotonically increasing, the optimal value is when When Monotonically decreasing, the optimal value is

[0183] S5.2, the solution of problem P3 is used as the user upload time;

[0184] S6, repeating steps S4 and S5 until the change in the user cost function value is less than a specified threshold;

[0185] In step S6, the formula for calculating the cost function value of the user equipment is: The cost function of the user equipment is defined as the weighted difference between the total energy consumption of the user equipment and the total conditional MI, that is, cost = E total -υMI sum , which reduces user energy consumption and improves perceived performance while ensuring task completion.

[0186] S7. Allocate resources to the user according to the user transmission power, local computing frequency and data upload time obtained in steps S3-S6. The user uses the allocated resources to unload tasks. The drone moves at the drone position obtained in step S2 and provides computing services to the user.

[0187] This embodiment considers that the drone is used as an edge computing node to provide edge computing services to users within the coverage area. The user uses the integrated wireless signal to offload the computing task to the drone in a partial offloading manner, and uses the integrated signal echo to perceive the surrounding environment. Considering the requirements of user data volume, perception performance, and drone location, the weighted mutual information of user energy consumption is used as the user's cost.

[0188] This embodiment first determines the position of the drone by using the Kmeans++ clustering classification method, and then decomposes the optimization problem into two sub-problems when the position of the drone is determined and solves them by an alternating iterative method.

[0189] This embodiment adopts the successive convex approximation method to obtain the user transmission power and the local calculation frequency, and obtains the relationship expression between the data upload time, the transmission power and the local calculation frequency through mathematical derivation. After the optimal values ​​of the user transmission power and the local calculation frequency are determined, the user data upload time can be obtained.

[0190] Example 2

[0191] This embodiment provides a device for joint optimization of drone position and resource allocation, including:

[0192] A detection module, used to detect the user's location;

[0193] A clustering module is used to cluster users based on user locations and a preset number of drone hovering time slots, and determine the drone location based on the clustering results;

[0194] A first calculation module is used to initialize the user data upload time according to the drone position, obtain the initial value of the upload time, and determine the user transmission power and the local calculation frequency based on the drone position and the initial value of the upload time;

[0195] The second calculation module is used to recalculate the user data upload time according to the drone position, the user transmission power and the local calculation frequency;

[0196] An iteration module, configured to repeatedly execute the first calculation module and the second calculation module until the change in the cost function value of the user equipment is less than a specified threshold;

[0197] The resource allocation module is used to allocate resources to users according to the obtained user transmission power, local computing frequency and user data upload time, so that users use the allocated resources to unload tasks, and the drone moves on the drone position obtained in the clustering module and provides computing services for users.

[0198] Example 3

[0199] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps of any method described in Embodiment 1 are implemented.

[0200] Example 4

[0201] This embodiment provides a computer device / equipment / system, including:

[0202] Memory, for storing computer programs / instructions;

[0203] A processor, configured to execute the computer program / instructions to implement the steps of any one of the methods described in Example 1.

[0204] Example 5

[0205] This embodiment provides a computer program product, including a computer program / instruction, which implements the steps of any method described in Embodiment 1 when executed by a processor.

[0206] 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.

[0207] Those skilled in the art will appreciate that the embodiments of the present disclosure may be provided as methods, systems or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present disclosure may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0208] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks 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 produce 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.

[0209] 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 comprising 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.

[0210] 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.

[0211] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure rather than to limit its protection scope. Although the present disclosure has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that after reading the present disclosure, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the disclosed claims to be approved.

Claims

1. A joint optimization method for drone location and resource allocation, characterized in that: include: S1. Detect user location; S2, clustering the users based on the user locations and the preset number of drone hovering time slots, and determining the drone locations based on the clustering results; S3. Initialize the user data upload time according to the drone position and obtain the initial value of the upload time; S4, determining the user transmission power and local calculation frequency based on the drone location and upload time; S5. Recalculate the user data upload time according to the drone position, user transmission power and local calculation frequency; S6, repeating steps S4 and S5 until the change in the cost function value of the user equipment is less than a specified threshold; S7, allocating resources to the user according to the user transmission power, local computing frequency and user data upload time obtained in steps S3-S6, so that the user uses the allocated resources to offload tasks, and the drone moves at the drone position obtained in step S2 and provides computing services to the user; In step S3, the user data upload time is initialized according to the drone position to obtain the initial value of the upload time, including: The hovering time slot number of the drone when user m communicates with the drone is S m , initialize the user data upload time to half of the hovering time slot length, and obtain the initial value of the upload time; The hovering time slot length is calculated using the following formula: in, is the length of the nth hovering time slot, is the total hovering time slot length; in, is the length of the nth flight slot, and the formula is: Among them, v is the speed of the drone when flying, L UAV (n) is the position of the UAV in hovering time slot n; In step S4, determining the user transmission power and the local calculation frequency based on the drone position and the upload time includes: S4.

1. Calculate the distance between the user device m and the drone in the hovering time slot n: Among them, L m is the position of user equipment m during the entire flight period; The channel gain between user equipment n and the drone in hovering time slot n is obtained as: Where ξ0 is the reference channel gain when the distance between the user and the drone is 1, and θ is the path loss factor; Calculate the total number of user devices communicating with the drone during the hovering time slot n as The available bandwidth of the user equipment communicating with the UAV in the hovering time slot n is The information transmission rate of user equipment m in the nth hovering time slot is obtained as: Where N0 is the power spectral density of the channel additive white Gaussian noise, P m (n) is the transmit power of user equipment m in hovering time slot n; S4.

2. When the drone is in a hovering state, the user equipment within the drone's communication range performs local computing and data upload at the same time. The drone's computing power is F. The drone distributes the drone's computing resources evenly to all users within the coverage area during the hovering time slot. When the drone is in a flying state, or the user equipment is not within the drone's communication range, the user equipment can only perform local computing. The local computing frequency of user equipment m in the nth hovering time slot is The local calculation frequency in the nth flight slot is Therefore, the amount of data calculated by user equipment m in the nth hovering time slot is Among them, D m represents the computational intensity of the task of user equipment m; the amount of data calculated in the nth flight time slot is The local computing energy consumption of user equipment m in the nth hovering time slot is The local computation energy consumption in the nth flight slot is where β loc Effective capacitance coefficient; S4.

3. Calculate the conditional mutual information of user equipment m in hovering time slot n as: Among them, s m (n) represents the waveform used by user equipment m to transmit data in hovering time slot n; φ m (n) represents the impulse response of the channel between user equipment m and the sensing target in hovering time slot n; z m (n) represents the reflected echo signal received by user equipment m in hovering time slot n; T s Indicates the duration of an OFDM symbol, Q m (n) represents the Fourier transform of the impulse response of the channel between the user device m and the sensing target in the hovering time slot n; S4.

4. Calculate the total energy consumption of all user devices in all time slots as: in, is the energy consumed by user equipment m in the nth hovering time slot, Indicates the length of time that user equipment m transmits data; The total conditional mutual information of the user equipment in all time slots is: S4.

5. Initialize the stopping condition of the successive convex approximation method, Taylor expansion point S4.

6. Based on the determined initial values ​​of the drone position and upload time, establish the user cost minimization problem P1, which is specifically expressed as follows: in, is the user device set, expressed as is the hover time slot index set, expressed as The flight slot index set is represented as S4.

7. Perform a first-order Taylor expansion on the left side of the constraint (C2) inequality of problem P1 to obtain the optimization problem P2, which is specifically expressed as: in S4.

8. Use the CVX toolbox to solve problem P2 and update the Taylor expansion point to the solution of problem P2; S4.9, repeat steps S4.7-S4.8 until the objective function value of problem P2 satisfies the stopping condition of the successive convex approximation method; S4.10, after the loop is completed, the solution to problem P2 is used as the user transmission power and the local calculation frequency; In step S5, the user data upload time is recalculated according to the drone position, the user transmission power and the local calculation frequency, including: S5.

1. Establish the optimization problem P3 based on the determined drone position, user transmission power and local computing frequency, which is specifically expressed as follows: When n=S m season f Taking partial derivatives we get: when When f Monotonically increasing, the optimal value is when When Monotonically decreasing, the optimal value is S5.2, taking the solution of problem P3 as the user data upload time; In step S6, the formula for calculating the cost function value of the user equipment is: cost = E total -υMI sum , where the cost function value cost of the user equipment is defined as the weighted difference between the total energy consumption of the user equipment and the total conditional MI.

2. The method for joint optimization of drone position and resource allocation according to claim 1, characterized in that: In step S2, clustering the users based on the user positions and the preset number of drone hovering time slots, and determining the drone positions according to the clustering results, including: S2.1, randomly select a user location as the first initial cluster center; S2.

2. Calculate the distance of all users from the currently determined nearest cluster center according to the formula Calculate the probability of all users’ positions becoming the next initial cluster center, where m represents the sequence number of user m; S2.

3. Determine the new cluster center by the roulette method. Divide the interval [0,1] into q intervals according to the probability of each user position becoming the next initial cluster center, where q is the total number of users. Then randomly generate a random number between 0 and 1. The user position corresponding to the interval where the random number is located is the new cluster center position. S2.4, repeat steps S2.2-S2.3 until the remaining N-1 initial cluster centers are selected, where N is the number of drone hovering time slots; S2.5, grouping users whose nearest cluster center is the same cluster center into the same cluster; S2.

6. The average value of the horizontal coordinates of all user positions in each cluster is used as the horizontal coordinate of the new cluster center, and the average value of the vertical coordinates of all user positions is used as the vertical coordinate of the new cluster center; S2.7, repeat steps S2.5-S2.6 until the change of all cluster centers is less than the threshold or the maximum number of cycles is reached; S2.

8. The hovering height of the drone is obtained as the maximum value of the lowest height covering all user devices in the current cluster and the lowest height of the drone; S2.9, the first position where the drone hovers is the position of the cluster center closest to the coordinate origin, and the cluster center position closest to the current drone position in the untraversed cluster center set is the next hovering position; S2.

10. Get the final position of the UAV in each hovering time slot.

3. A device for joint optimization of drone position and resource allocation, characterized in that: include: A detection module, used to detect the user's location; A clustering module is used to cluster users based on user locations and a preset number of drone hovering time slots, and determine the drone location based on the clustering results; A first calculation module is used to initialize the user data upload time according to the drone position, obtain the initial value of the upload time, and determine the user transmission power and the local calculation frequency based on the drone position and the initial value of the upload time; The second calculation module is used to recalculate the user data upload time according to the drone position, the user transmission power and the local calculation frequency; An iteration module, configured to repeatedly execute the first calculation module and the second calculation module until the change in the cost function value of the user equipment is less than a specified threshold; A resource allocation module is used to allocate resources to users according to the obtained user transmission power, local computing frequency and user data upload time, so that users use the allocated resources to unload tasks, and the drone moves on the drone position obtained in the clustering module and provides computing services to the users; Among them, in the first calculation module, the user data upload time is initialized according to the drone position to obtain the initial value of the upload time, including: The hovering time slot number of the drone when user m communicates with the drone is S m , initialize the user data upload time to half of the hovering time slot length, and obtain the initial value of the upload time; The hovering time slot length is calculated using the following formula: in, is the length of the nth hovering time slot, is the total hovering time slot length; in, is the length of the nth flight slot, and the formula is: Among them, v is the speed of the drone when flying, L UAV (n) is the position of the UAV in hovering time slot n; In step S4, determining the user transmission power and the local calculation frequency based on the drone position and the upload time includes: S4.

1. Calculate the distance between the user device m and the drone in the hovering time slot n: Among them, L m is the position of user equipment m during the entire flight period; The channel gain between user equipment n and the drone in hovering time slot n is obtained as: Where ξ0 is the reference channel gain when the distance between the user and the drone is 1, and θ is the path loss factor; Calculate the total number of user devices communicating with the drone during the hovering time slot n as The available bandwidth of the user equipment communicating with the UAV in the hovering time slot n is The information transmission rate of user equipment m in the nth hovering time slot is obtained as: Where N0 is the power spectral density of the channel additive white Gaussian noise, P m (n) is the transmit power of user equipment m in hovering time slot n; S4.

2. When the drone is in a hovering state, the user equipment within the drone's communication range performs local computing and data upload at the same time. The drone's computing power is F. The drone distributes the drone's computing resources evenly to all users within the coverage area during the hovering time slot. When the drone is in a flying state, or the user equipment is not within the drone's communication range, the user equipment can only perform local computing. The local computing frequency of user equipment m in the nth hovering time slot is The local calculation frequency in the nth flight slot is Therefore, the amount of data calculated by user equipment m in the nth hovering time slot is Among them, D m represents the computational intensity of the task of user equipment m; the amount of data calculated in the nth flight time slot is The local computing energy consumption of user equipment m in the nth hovering time slot is The local computation energy consumption in the nth flight time slot is where β loc Effective capacitance coefficient; S4.

3. Calculate the conditional mutual information of user equipment m in hovering time slot n as: Among them, s m (n) represents the waveform used by user equipment m to transmit data in hovering time slot n; φ m (n) represents the impulse response of the channel between user equipment m and the sensing target in hovering time slot n; z m (n) represents the reflected echo signal received by user equipment m in hovering time slot n; T s Indicates the duration of an OFDM symbol, Q m (n) represents the Fourier transform of the impulse response of the channel between the user device m and the sensing target in the hovering time slot n; S4.

4. Calculate the total energy consumption of all user devices in all time slots as: in, is the energy consumed by user equipment m in the nth hovering time slot, Indicates the length of time that user equipment m transmits data; The total conditional mutual information of the user equipment in all time slots is: S4.

5. Initialize the stopping condition of the successive convex approximation method, Taylor expansion point S4.

6. Based on the determined initial values ​​of the drone position and upload time, establish the user cost minimization problem P1, which is specifically expressed as follows: in, is the user device set, expressed as is the hover time slot index set, expressed as The flight slot index set is represented as S4.

7. Perform a first-order Taylor expansion on the left side of the constraint (C2) inequality of problem P1 to obtain the optimization problem P2, which is specifically expressed as: in S4.

8. Use the CVX toolbox to solve problem P2 and update the Taylor expansion point to the solution of problem P2; S4.9, repeat steps S4.7-S4.8 until the objective function value of problem P2 satisfies the stopping condition of the successive convex approximation method; S4.10, after the loop is completed, the solution to problem P2 is used as the user transmission power and the local calculation frequency; In the second calculation module, the user data upload time is recalculated according to the drone position, the user transmission power and the local calculation frequency, including: S5.

1. Establish the optimization problem P3 based on the determined drone position, user transmission power and local computing frequency, which is specifically expressed as follows: When n=S m season f Taking partial derivatives we get: when When f Monotonically increasing, the optimal value is when When Monotonically decreasing, the optimal value is S5.2, taking the solution of problem P3 as the user data upload time; In the iteration module, the formula for calculating the cost function value of the user equipment is: cost = E total -υMI sum , where the cost function value cost of the user equipment is defined as the weighted difference between the total energy consumption of the user equipment and the total conditional MI.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method described in any one of claims 1 to 2 are implemented.

5. A computer device / equipment / system, characterized in that: include: Memory, for storing computer programs / instructions; A processor, configured to execute the computer program / instructions to implement the steps of the method according to any one of claims 1 to 2.

6. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 2 are implemented.