High-altitude platform task offloading method based on unmanned aerial vehicle relay

Abstract

The application discloses a high-altitude platform task unloading method based on unmanned aerial vehicle relaying, wherein each user sends a task unloading demand to an unloading server, each high-altitude platform and each unmanned aerial vehicle sends state information to the unloading server, the unloading server adopts a preset resource allocation method, allocates high-altitude platform resources according to the task unloading demand of each user and the state information of each high-altitude platform, obtains a task allocation variable, then carries out reverse auction on the unmanned aerial vehicles, obtains a relay allocation variable and a final winner set, and finally carries out task unloading on each user in the final winner set according to the relay allocation variable and the task allocation variable. The application is aimed at a high-altitude computing system for building an aerial relay link through unmanned aerial vehicles, and realizes efficient allocation of high-altitude platform and unmanned aerial vehicle resources.

Description

Technical Field

[0001] This invention belongs to the field of aerial platform technology, and more specifically, relates to an aerial platform task unloading method based on hybrid auction. Background Technology

[0002] A high-altitude platform is an aircraft or airship system that can be deployed long-term in the stratosphere (typically at an altitude of 20–25 kilometers). A high-altitude platform equipped with an edge server refers to a platform that integrates high-performance computing units, enabling it to perform edge data processing and service distribution in near-air regions. This allows it to be rapidly deployed to target airspace according to mission requirements, providing flexible and mobile computing support for ground or air users.

[0003] While cloud computing provides centralized, large-scale computing power, its data centers are typically located far from end users. Due to network transmission distances and intermediate nodes, there is high communication latency, making it difficult to meet the real-time processing requirements of latency-sensitive tasks. Edge computing, by moving computing power to the network edge, can perform local processing close to the data source, reducing latency. However, it relies on fixed ground infrastructure, and edge servers need to be pre-deployed in specific areas, making it impossible to dynamically adjust service location and coverage according to changes in task scenarios.

[0004] In contrast, airborne computing platforms such as High Altitude Platforms (HAPs) and Unmanned Aerial Vehicles (UAVs) possess high mobility and rapid deployment capabilities, enabling them to respond to temporary computing power needs in specific scenarios (such as disaster areas, remote regions, or large event venues), achieving flexible and dynamic service coverage, thereby expanding the boundaries of computing power services in both spatial and temporal dimensions. However, in existing HAP systems, the efficient offloading of ground user tasks faces several key challenges: how to execute offloading decisions, how to select relay UAVs, and which users are eligible for offloading, all of which require further research. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a high-altitude platform task offloading method based on UAV relay, which is aimed at high-altitude computing systems that build air relay links through UAVs, and realizes efficient allocation of high-altitude platform and UAV resources.

[0006] To achieve the above-mentioned objectives, the present invention provides a high-altitude platform task offloading method based on UAV relay, comprising the following steps:

[0007] S1: Get the user set Each user Unload its task requirements Send to the uninstallation server, where, , Indicates user The system efficiency of offloading tasks to remote operation on a high-altitude platform. Indicates user The resource demand vector, Indicates user Tasks on resources The amount of resources required, , This represents the set of resource types that the user task needs to use. Indicates user The amount of task data, Indicates user Task transmission latency limits;

[0008] Get a collection of high-altitude platforms Each high-altitude platform Its status information Send to the uninstallation server, where, , Indicates high-altitude platform The unit system overhead of performing the uninstallation task. Represents a resource provision vector. Indicates high-altitude platform Resources The amount provided;

[0009] Get drone collection Each drone Its status information Send to the uninstallation server, where, , Indicates drone The unit system overhead for relaying, Indicates user With drones transmission rate Indicates drone With aerial platform The transmission rate;

[0010] S2: The unloading server uses a preset resource allocation method, based on each user. Task uninstallation requirements and each aerial platform Status information Perform high-altitude platform resource allocation to obtain task allocation variables. ,when Indicates that the user The task was unloaded onto the high-altitude platform. Remote operation, when Indicates not to use users The task was unloaded onto the high-altitude platform. Remote operation, thereby obtaining the winner set of high-altitude platform resource allocation. ;

[0011] S3: The offloading server performs a reverse auction for the drone, obtaining the user's... The mission was carried out by drones Relay allocation variables unloaded to remote operation on an aerial platform ,when Indicates that the user The mission was carried out by drones Unloaded to a high-altitude platform for remote operation, when Indicates not to use users The mission was carried out by drones The specific steps for unloading and remotely running the software on an aerial platform are as follows:

[0012] S3.1: Gather the winners Users in the middle are subject to task transmission latency limits Sort the results from smallest to largest to obtain the winner sequence. ;

[0013] S3.2: Initialize all relay assignment variables ;

[0014] S3.3: Assemble the drones The drones in the sequence are sorted according to their unit system overhead from smallest to largest to obtain the drone sequence. ;

[0015] S3.4: Initialize the final winner set ;

[0016] S3.5: From the winners' sequence Select the first user The high-altitude platform assigned to it is recorded as and remove it from the winners' sequence. Delete;

[0017] S3.6: Select the drone sequence in sequence Each drone Determine whether the conditions are met simultaneously. and Once there are drones If the constraints are met, proceed to step S3.7; if none of the drones meet the conditions, proceed to step S3.8.

[0018] S3.7: Let the relay assignment variable and order Proceed to step S3.9;

[0019] S3.8: Update task assignment variables Proceed to step S3.9;

[0020] S3.9: Determine if it is a winner sequence If yes, the drone reverse auction ends; otherwise, return to step S3.5.

[0021] S4: The offloading server allocates variables based on the relay. and task allocation variables For the final victors Various users Uninstall: When and At that time, through drones As a relay user The task was unloaded onto the high-altitude platform. Execute.

[0022] This invention relates to a high-altitude platform task unloading method based on UAV relay. Each user sends its task unloading request to the unloading server, and each high-altitude platform and each UAV sends its status information to the unloading server. The unloading server uses a preset resource allocation method to allocate high-altitude platform resources according to each user's task unloading request and each high-altitude platform's status information, obtaining task allocation variables. Then, a reverse auction is conducted on the UAVs to obtain relay allocation variables and a final set of winners. Finally, task unloading is performed on each user in the final set of winners according to the relay allocation variables and task allocation variables.

[0023] This invention addresses the UAV-assisted computing scenario in high-altitude computing systems, solving the resource allocation problem of high-altitude platforms and UAVs within these systems. It effectively improves the resource utilization rate of high-altitude resource platforms, meets the reliability requirements, and has practical application value. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the high-altitude platform task unloading based on UAV relay in this invention;

[0025] Figure 2 This is a flowchart illustrating a specific implementation method of the high-altitude platform task unloading method based on UAV relay of the present invention;

[0026] Figure 3 This is a flowchart of the one-way auction of high-altitude platform resources in this embodiment;

[0027] Figure 4 This is a flowchart of the reverse auction of drone resources in this invention;

[0028] Figure 5This is a flowchart of the optimization of the final winner set based on system performance gains in this embodiment. Detailed Implementation

[0029] The specific embodiments of the present invention will now be described with reference to the accompanying drawings to enable those skilled in the art to better understand the invention. It should be particularly noted that in the following description, detailed descriptions of known functions and designs that might obscure the main content of the invention will be omitted here.

[0030] To better illustrate the technical solution of the present invention, the principle of the present invention will be briefly explained first. Figure 1 This is a schematic diagram illustrating the high-altitude platform task unloading based on UAV relay according to the present invention. Figure 1 As shown, User 1's task is relayed to High-Altitude Platform 1 via Drone 1 for remote execution, while User 2's task is relayed to High-Altitude Platform 2 via Drone 2 for remote execution. Generally, this process involves four steps: First, acquiring information about the users, the High-Altitude Platform, and the Drones; second, a one-way auction of High-Altitude Platform resources; third, a reverse auction of Drone resources; and fourth, task unloading.

[0031] Based on analysis, this invention models the unloading of high-altitude platform tasks based on UAV relay as a linear programming problem, which is expressed as follows:

[0032] (1)

[0033] st (1a)

[0034] (1b)

[0035] (1c)

[0036] (1d)

[0037] (1e)

[0038] in, Represents a set of users. This represents a collection of high-altitude platforms. Indicates a collection of drones. Indicates the assignment variable, Indicates user The mission was carried out by drones Unload to an aerial platform for remote execution. Define allocation variables. , Indicates user The task was unloaded onto the high-altitude platform. Execute remotely. Indicates user The system efficiency of offloading tasks to remote operation on a high-altitude platform. Indicates user The amount of task data, , These represent high-altitude platforms. drones The unit system overhead of performing the uninstallation task.

[0039] Objective condition (1) represents maximizing system efficiency; constraint condition (1a) states that the amount of resources allocated to each high-altitude platform cannot exceed its available resources; constraint condition (1b) states that user tasks must meet latency constraints when executed remotely; constraint condition (1c) states that user tasks can only be transferred via one UAV when offloading to a high-altitude platform; constraint condition (1d) states that user tasks can be offloaded to at most one high-altitude platform; constraint condition (1e) represents the allocation variable. and It can only take the value 1 or 0.

[0040] Based on the above analysis, this invention proposes a high-altitude platform task unloading method based on UAV relay. Figure 2 This is a flowchart illustrating a specific implementation method of the high-altitude platform mission unloading method based on UAV relay of the present invention. Figure 2 As shown, the specific steps of the high-altitude platform mission unloading method based on UAV relay of the present invention include:

[0041] S201: Obtain information about users, high-altitude platforms, and drones:

[0042] Get user set Each user Unload its task requirements Send to the uninstallation server, where, , Indicates user The system efficiency of offloading tasks to remote operation on a high-altitude platform. Indicates user The resource demand vector, Indicates user Tasks on resources The amount of resources required, , This represents the set of resource types that the user task needs to use. Resource types typically include CPU and memory. Indicates user The amount of task data, Indicates user The task transmission latency limit, that is, the task must be within the latency limit. The data is transmitted internally to the high-altitude platform.

[0043] Get a collection of high-altitude platforms Each high-altitude platform Its status information Send to the uninstallation server, where, , Indicates high-altitude platform The unit system overhead of performing an unload task is the system overhead required to provide one unit of resource. Represents a resource provision vector. Indicates high-altitude platform Resources The amount provided.

[0044] Get drone collection Each drone Its status information Send to the uninstallation server, where, , Indicates drone The unit system overhead for relaying, Indicates user With drones transmission rate Indicates drone With aerial platform The transmission rate. In practical applications, the network transmission rate of a drone is related to communication technology, signal strength and quality, spectrum and bandwidth, environmental interference, and hardware performance. An empirical value can be set according to actual needs.

[0045] S202: Allocation of high-altitude platform resources:

[0046] The unloading server uses a preset resource allocation method, based on each user's task unloading requirements. and each aerial platform Status information Perform high-altitude platform resource allocation to obtain task allocation variables. ,when Indicates that the user The task was unloaded onto the high-altitude platform. Remote operation, when Indicates not to use users The task was unloaded onto the high-altitude platform. Remote operation, thereby obtaining the winner set of high-altitude platform resource allocation. .

[0047] Currently, auction mechanisms are commonly used to solve the problem of distributed resource allocation. In this embodiment, a one-way auction is used to allocate resources for the high-altitude platform. Figure 3 This is a flowchart of the one-way auction of high-altitude platform resources in this embodiment. For example... Figure 3 As shown, the specific steps for unidirectional auction of high-altitude platform resources in this embodiment include:

[0048] S301: Calculate edge contribution value:

[0049] The edge contribution value is the difference between a user's system performance and the system overhead of remote execution on a high-altitude platform. A higher edge contribution value indicates a greater increase in system performance, making it a priority to offload the user's tasks to this high-altitude platform. Therefore, in this embodiment's one-way auction, the following formula is first used to calculate the user's... The task was unloaded onto the high-altitude platform. Edge contribution value of remote operation :

[0050] .

[0051] S302: Initialize the pairing set and assign variables:

[0052] Each user is paired with all aerial platforms to obtain an initial pairing set. Then initialize all task assignment variables. .

[0053] S303: User Screening - High-Altitude Platform Matching:

[0054] From the pairing set The user with the highest marginal contribution value was selected from the pool and paired with the high-altitude platform. and remove it from the pairing set Delete it.

[0055] S304: Determine whether the available resources of the aerial platform meet the user's needs, i.e., determine whether... ,in This indicates the preset overselling rate. If the target is met, proceed to step S305; otherwise, proceed to step S306.

[0056] Because this invention auctions off high-altitude platform resources first, and then drone resources, it's possible that a user might be allocated high-altitude platform resources, but the drone's transmission rate cannot meet the user's latency requirements, thus preventing the task from being unloaded. Therefore, an oversold rate is set. The system first allocates excess resources from the high-altitude platform to users, and then determines which users will be the winners when auctioning off drone resources, thus avoiding resource waste.

[0057] S305: Assign unloading tasks:

[0058] When resource conditions are met, it indicates that the aerial platform The available resources can meet the user's needs. The resource requirements then cause the task allocation variable to... , indicating user The task was to unload the equipment onto a high-altitude platform. And because of users Once an assignment has been made, there's no need to consider other pairings; therefore, it's necessary to assign users... Other pairings from the pairing set Exclusion, i.e., updating the pairing set. Then proceed to step S306.

[0059] S306: Determine if a set is paired If yes, it means all pairs have been checked and the allocation is complete; otherwise, return to step S303.

[0060] Because the resources of the high-altitude platform in this embodiment are over-allocated, and it cannot be guaranteed that the task latency of each user will meet the requirements, therefore, a winner set is determined. Not all users are the ultimate winners.

[0061] S203: Reverse Auction of Drone Resources:

[0062] The next step involves a reverse auction of the drone, yielding the user's [property / resource]. The mission was carried out by drones Relay allocation variables unloaded to remote operation on an aerial platform ,when Indicates that the user The mission was carried out by drones Unloaded to a high-altitude platform for remote operation, when Indicates not to use users The mission was carried out by drones The drones are offloaded to a high-altitude platform for remote operation, thereby determining which drones relay missions for which users. Figure 4 This is a flowchart of the reverse auction of drone resources in this invention. For example... Figure 4 As shown, the specific steps of reverse auctioning drone resources in this invention include:

[0063] S401: Winners' Order:

[0064] Gather the winners Users in the middle are subject to task transmission latency limits Sort the results from smallest to largest to obtain the winner sequence. .

[0065] S402: Initialize relay allocation variables:

[0066] Initialize all relay allocation variables .

[0067] S403: Drone Sequencing:

[0068] A collection of drones The drones in the sequence are sorted according to their unit system overhead from smallest to largest to obtain the drone sequence. The purpose of drone sorting is to prioritize drones with lower unit costs for relaying, in order to meet the latency requirements of user tasks as much as possible.

[0069] S404: Initialize the final winner set .

[0070] S405: Select User:

[0071] From the winners' sequence Select the first user The high-altitude platform assigned to it is recorded as Then remove it from the winner sequence. Delete it.

[0072] S406: Screening Drone Relays:

[0073] Selecting drone sequences in sequence Each drone In other words, we should start by considering drones with lower unit system overhead and determine whether they simultaneously meet the requirements. and Once there are drones If the constraints are met, proceed to step S407; if none of the drones meet the conditions, proceed to step S408.

[0074] S407: Assigning drone relays:

[0075] Let relay assignment variable and will users Join the final winners' group , that is to say Proceed to step S409.

[0076] S408: Reclaim task assignment:

[0077] If all drones fail to meet latency constraints, or the high-altitude platform cannot meet resource requirements, then the current user... The task cannot be unloaded to the high-altitude platform. Update task assignment variables Proceed to step S409.

[0078] S409: Determine if it is a winner's sequence If yes, the drone reverse auction ends; otherwise, return to step S405.

[0079] In practical applications, system performance gains also need to be considered when unloading tasks, and the final winner set should be determined accordingly. Further optimization is needed. Figure 5 This is a flowchart illustrating the optimization of the final winner set based on system performance gains in this embodiment. For example... Figure 5 As shown, the specific steps for optimizing the final winner set based on system performance gains in this embodiment include:

[0080] S501: Calculate preliminary system overhead:

[0081] Traverse the final winner set Each user in The following method is used to calculate the preliminary system overhead considering only the allocation of resources on the high-altitude platform. :

[0082] From user set Delete user Then, the high-altitude platform resources are simulated and allocated again. During this process, whenever a user receives a high-altitude platform resource allocation, it is determined whether the remaining high-altitude platform resources can meet the user's needs. If the resource demand is met, the remaining users will continue to be allocated resources; otherwise, the user who most recently received a high-altitude resource allocation will be selected as the new user. Key users .

[0083] Calculate users Preliminary system overhead ,in In the simulated allocation of high-altitude platform resources, the user is considered a key user. High-altitude platform for resource allocation Indicates based on the set of winners What is obtained is for the user High-altitude platform for resource allocation The price coefficient representing marginal contribution.

[0084] S502: Determine the system overhead of key units of the UAV:

[0085] From drone sequence Select drones that meet the following conditions Drones The previous drone provided relay services to the user. None of the drones thereafter provided relay services to the user. If it exists, then the system overhead of the key unit of the drone will be... If the drone If it does not exist, it will increase the overhead of the drone's critical unit system. .

[0086] S503: Calculate the final system overhead:

[0087] For the final victors Each user in The final system overhead is calculated using the following formula. :

[0088] ,

[0089] in, Indicates user The amount of task data.

[0090] S504: Selecting the Winner

[0091] For the final victors Each user in Determine whether , Indicates user The system performance of offloading tasks to a remotely running high-altitude platform; if so, it indicates that the user... The current resource allocation scheme offers system performance benefits; therefore, retain this scheme and take no further action. Otherwise, it indicates that the user... The system overhead of offloading the mission to an aerial platform exceeded system efficiency, leading to the final winners' set. Lieutenant General User Delete, Update , .

[0092] S204: Task Unloading:

[0093] The offloading server allocates variables based on the relay. and task allocation variables For the final victors Various users Uninstall: When and At that time, through drones As a relay user The task was unloaded onto the high-altitude platform. Execute.

[0094] To illustrate the implementation process of this invention, a specific example is used to experimentally verify the invention. This embodiment assumes four users, one high-altitude platform, and two drones. Two types of resources are assumed to be provided. Table 1 is a list of resource requirements for the users in this embodiment.

[0095]

[0096] Table 1

[0097] In this embodiment, the unit system overhead of the high-altitude platform is 0.1, and the resource provision is... and Table 2 is a list of drone and user status information in this embodiment.

[0098]

[0099] Table 2

[0100] First, a one-way auction of high-altitude platform resources is conducted, with an oversold rate set. Therefore, in the oversold resources, the high-altitude platform can allocate 12 × 1.3 = 15.6 and 24 × 1.3 = 31.2 resources.

[0101] Calculate the edge contribution value for each user:

[0102] ,

[0103] ,

[0104] ,

[0105] .

[0106] Constructing the initial user-high-altitude platform pairing set ={(1,1),(2,1),(3,1),(4,1)}.

[0107] Select the pair (1,1) with the largest edge contribution value from the pairing set and update the pairing set. ={(2,1),(3,1),(4,1)}. Since the over-sold resources of the high-altitude platform can meet the resource needs of User 1, let the task allocation variable... .

[0108] Continue by selecting the pair (2,1) with the largest edge contribution from the pairing set and updating the pairing set. ={(3,1),(4,1)}. Since the over-sold resources of the high-altitude platform can meet the resource needs of User 2, the task allocation variable is set to {(3,1),(4,1)}. .

[0109] Continue by selecting the pair (3,1) with the largest edge contribution from the pairing set and updating the pairing set. ={(4,1)}. Since the over-sold resources of the high-altitude platform can meet the resource needs of user 3, let the task allocation variable be {(4,1)}. .

[0110] Finally, select the pair with the largest edge contribution (4,1) from the pairing set and update the pairing set. Because the over-sold resources on the high-altitude platform cannot meet the resource needs of user 4, the task allocation variable is changed. .

[0111] Thus, a group of winners was formed. .

[0112] Next, the drone resources will be auctioned off in reverse. First, the drones will be sorted from smallest to largest according to their unit cost to obtain an inorganic sequence. Select the winners' group. User 1 in the list is removed to obtain the winner set. Selecting UAV sequences The first drone 1 in the system, since its transmission time as a relay is 0.45s, is less than the transmission latency of user 1, thus meeting user 1's transmission latency requirements. Furthermore, the available resources of the high-altitude platform can also meet user 1's needs. Therefore, the user-drone allocation variable is set to... To bring together the final victors .

[0113] Then select the winners' set. User 2 in the list is removed to obtain the winner set. Selecting UAV sequences The first drone 1 in the system, since its transmission time as a relay is 0.44s, is less than the transmission latency of user 2, thus meeting user 2's transmission latency requirements. Furthermore, the available resources of the high-altitude platform can also meet user 2's needs. Let the user-drone allocation variable be... To bring together the final victors .

[0114] Select the winners set User 3 in the list is removed to obtain the winner set. Selecting UAV sequences The first drone, 1, has a transmission time of 0.48 seconds as a relay, which meets User 3's transmission latency requirements. However, the available resources of the high-altitude platform cannot meet User 3's needs. Therefore, a new drone sequence will be selected. The second drone, 2, although its transmission time as a relay is 0.5 seconds, which meets User 3's transmission latency requirements, cannot meet User 3's needs due to insufficient available resources on the high-altitude platform. Therefore, User 3 is the loser, and the final winners are... .

[0115] Next, the set of final winners will be determined based on system performance gains. Optimization is performed. First, the initial system overhead for User 1 is calculated. Since the user selection order is 2, 3, and 4 when User 1 is not involved in allocation, and the over-sold resources on the high-altitude platform cannot meet User 1's requirements when User 4 acquires resources, User 4 is a critical user for User 1. The initial system overhead for User 1 is then calculated. :

[0116] .

[0117] Then, calculate the initial system overhead for User 2. Since the user selection order is 1, 3, and 4 when User 2 is not involved in allocation, and the oversold resources on the high-altitude platform cannot meet User 2's requirements when User 4 acquires resources, User 4 is a critical user for User 2. Calculate the initial system overhead for User 2. :

[0118] .

[0119] From drone sequence If key UAV 2 is selected from the list, then the key unit system overhead of the UAV is determined. Then calculate the final system overhead for user 1 and user 2:

[0120] ,

[0121] .

[0122] because Therefore, the resource allocation scheme for User 1 is retained. Meanwhile... Therefore, the resource allocation scheme of User 2 is retained, and the final winners are... constant.

[0123] Although the illustrative specific embodiments of the present invention have been described above to enable those skilled in the art to understand the invention, it should be understood that the invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the invention as defined and determined by the appended claims, and all inventions utilizing the concept of the present invention are protected.

Claims

1. A method for unloading missions on a high-altitude platform based on UAV relay, characterized in that, Includes the following steps: S1: Get the user set Each user Unload its task requirements Send to the uninstallation server, where, , Indicates user The system efficiency of offloading tasks to remote operation on a high-altitude platform. Indicates user The resource demand vector, Indicates user Tasks on resources The amount of resources required, , This represents the set of resource types that the user task needs to use. Indicates user The amount of task data, Indicates user Task transmission latency limits; Get a collection of high-altitude platforms Each high-altitude platform Its status information Send to the uninstallation server, where, , Indicates high-altitude platform The unit system overhead of performing the uninstallation task. Represents a resource provision vector. Indicates high-altitude platform Resources The amount provided; Get drone collection Each drone Its status information Send to the uninstallation server, where, , Indicates drone The unit system overhead for relaying, Indicates user With drones transmission rate Indicates drone With aerial platform The transmission rate; S2: The unloading server uses a preset resource allocation method, based on each user. Task uninstallation requirements and each aerial platform Status information Perform high-altitude platform resource allocation to obtain task allocation variables. ,when Indicates that the user The task was unloaded onto the high-altitude platform. Remote operation, when Indicates not to use users The task was unloaded onto the high-altitude platform. Remote operation, thereby obtaining the winner set of high-altitude platform resource allocation. ; S3: The offloading server performs a reverse auction for the drone, obtaining the user's... The mission was carried out by drones Relay allocation variables unloaded to remote operation on an aerial platform ,when Indicates that the user The mission was carried out by drones Unloaded to a high-altitude platform for remote operation, when Indicates not to use users The mission was carried out by drones The specific steps for unloading and remotely running the software on an aerial platform are as follows: S3.1: Gather the winners Users in the middle are subject to task transmission latency limits Sort the results from smallest to largest to obtain the winner sequence. ; S3.2: Initialize all relay assignment variables ; S3.3: Assemble the drones The drones in the sequence are sorted according to their unit system overhead from smallest to largest to obtain the drone sequence. ; S3.4: Initialize the final winner set ; S3.5: From the winners' sequence Select the first user The high-altitude platform assigned to it is recorded as and remove it from the winners' sequence. Delete; S3.6: Select the UAV sequence in sequence Each drone Determine whether the conditions are met simultaneously. and Once there are drones If the constraints are met, proceed to step S3.7; if none of the drones meet the conditions, proceed to step S3.

8. S3.7: Let the relay assignment variable and order Proceed to step S3.9; S3.8: Update task assignment variables Proceed to step S3.9; S3.9: Determine if it is a winner sequence If yes, the drone reverse auction ends; otherwise, return to step S3.

5. S4: The offloading server allocates variables based on the relay. and task allocation variables For the final victors Various users Uninstall: When and At that time, through drones As a relay user The task was unloaded onto the high-altitude platform. Execute.

2. The high-altitude platform task unloading method according to claim 1, characterized in that, The specific steps for allocating high-altitude platform resources in step S2 include: S2.1: The following formula is used to calculate the user's... The task was unloaded onto the high-altitude platform. Edge contribution value of remote operation : ； S2.2: Pair each user with all high-altitude platforms to obtain an initial pairing set. Then initialize all task assignment variables. ; S2.3: From the pairing set The user with the highest marginal contribution value was selected from the pool and paired with the high-altitude platform. and remove it from the pairing set Delete; S2.4: Determine if the condition is met. ,in This indicates the preset overselling rate. If the target is met, proceed to step S2.5; otherwise, proceed to step S2.

6. S2.5: Let the task assignment variable... and update the pairing set. Then proceed to step S2.6; S2.6: Determine if sets are paired If yes, the allocation ends and proceeds to step S2.7; otherwise, return to step S2.

3.

3. The high-altitude platform task unloading method according to claim 1, characterized in that, Step S3 also includes evaluating the final winner set based on system performance gains. The optimization process is as follows: S3.10.1: Traverse the set of final winners Each user in The following method is used to calculate the preliminary system overhead considering only the allocation of resources on the high-altitude platform. : From user set Delete user Then, the high-altitude platform resources are simulated and allocated again. During this process, whenever a user receives a high-altitude platform resource allocation, it is determined whether the remaining high-altitude platform resources can meet the user's needs. If the resource demand is met, the remaining users will continue to be allocated resources; otherwise, the user who most recently received a high-altitude resource allocation will be selected as the new user. Key users ; Calculate users Preliminary system overhead ,in In the simulated allocation of high-altitude platform resources, the user is considered a key user. High-altitude platform for resource allocation Indicates based on the set of winners What is obtained is for the user A high-altitude platform for resource allocation; S3.10.2: From the drone sequence Select drones that meet the following conditions Drones The previous drone provided relay services to the user. None of the drones thereafter provided relay services to the user. If it exists, then the system overhead of the key unit of the drone will be... If the drone If it does not exist, it will increase the overhead of the drone's critical unit system. ; S3.10.3: Set of final winners Each user in The final system overhead is calculated using the following formula. : ， in, Indicates user The amount of task data; S3.10.4: Set of final winners Each user in Determine whether , Indicates user The task is offloaded to the remotely running system on the high-altitude platform. If so, no action is taken; otherwise, the final winner is selected. Lieutenant General User Delete, Update , .

Images