Near-earth satellite communication task unloading method based on ground cooperative computing and program product

Abstract

The invention discloses a near-earth satellite communication task unloading method based on ground cooperative computing and a program product, a satellite communication task unloading model is established, and the satellite communication task unloading model aims at the minimum task unloading total time delay. A task unloading rate strategy of a satellite, an unloading path strategy of a task on the satellite and an unloading path strategy of the task on a ground station are used as variables to be solved; the task unloading total time delay is specifically the sum of the calculation time delay of the task on the satellite, the transmission time delay of the task unloading from the satellite to the ground station, the transmission time delay of the task between the ground stations and the calculation time delay of the task on the ground station, and then solving the satellite communication task unloading model to obtain a task unloading strategy. According to the method, the unloading strategy is relatively small in time delay and high in robustness, the defect that the LEO calculation capability is limited is effectively overcome, the upper limit of the task quantity accepted by a single LEO is greatly increased, and meanwhile, the problem of non-uniform task processing distribution of a ground station is solved.

Description

Technical Field

[0001] The present invention relates to satellite mission offloading technology, and in particular to a near-earth satellite communication mission offloading method and program product based on ground collaborative computing. Background Art

[0002] With the development of 5G and 6G communication technologies, the defects of traditional ground communication systems restricted by geographical factors have become increasingly obvious, and it is difficult to meet people's communication needs. In contrast, satellite communication networks have a wider coverage, and service access is simple and fast. Even in remote areas where ground signal stations are scarce, satellites can significantly improve communication efficiency. Due to the special geographical location of satellites, they are difficult to be interfered by ground environmental factors. When serious natural disasters occur on the ground, reliable communication services can still be guaranteed. Therefore, satellite communication has become a hot topic in today's communication technology research.

[0003] In terms of satellite usage, satellites can be divided into GEO geostationary Earth Orbit Satellite, MEO medium Earth Orbit Satellite, and LEO low Earth Orbit Satellite according to their orbits. Although GEO satellites were widely used in the early days of satellite communications due to their large coverage brought by their high orbits, with the growing business needs, the higher satellite orbits of GEO and MEO and the data transmission delays caused by long-distance communications cannot meet the demand for low-latency satellite communications. In comparison, LEO's low orbit, low manufacturing cost, and large quantity are more able to meet current needs.

[0004] However, in terms of satellite processing tasks, the low-cost and small LEO units bring about insufficient computing power, and it is difficult to meet the needs of a large number of computing tasks by relying solely on their own computing power. Several solutions currently focused on include: ground collaborative computing, satellite constellation collaborative computing, satellite edge computing, and optimized satellite algorithms. Among them, the advantages of ground collaborative computing, such as strong ground station computing power, low latency and high bandwidth, make ground collaborative computing have a higher upper limit in optimizing satellite processing task delays and increasing the number of tasks accepted by satellites.

[0005] In terms of downloading ground collaborative computing tasks, various solutions are currently adopted. For the correspondence between satellites and ground stations, there are single satellite corresponding to single ground station, multiple satellites corresponding to single ground station, single satellite corresponding to multiple ground stations, and multiple satellites corresponding to multiple ground stations. In terms of the multiple access schemes adopted, there are multiple forms of multiple access schemes such as frequency division multiple access, time division multiple access, and code division multiple access. However, the current task offloading latency still needs to be further optimized. Summary of the invention

[0006] In view of the problems existing in the prior art, the purpose of the present invention is to provide a near-Earth satellite communication task offloading method and program product based on ground collaborative computing with low latency.

[0007] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions: A method for offloading near-Earth satellite communication tasks based on ground collaborative computing comprises the following steps: (1) Establishing a satellite communication task offloading model, wherein the satellite communication task offloading model aims to minimize the total delay of task offloading and takes the task offloading rate strategy of the satellite as the α , Task offloading path strategy on satellite c , Task offloading path strategy on the ground station β is the variable to be solved, and the total delay of task offloading is specifically the sum of the calculation delay of the task on the satellite, the transmission delay of the task offloading from the satellite to the ground station, the transmission delay of the task between the ground stations and the calculation delay of the task on the ground station; (2) Set the number of outer loop iterations a =0, the number of inner loop iterations b=0; (3) a The satellite task offloading rate strategy during the bth outer cycle and the bth inner cycle α (a)(b) , Task offloading path strategy on satellite c (a)(b) , Task offloading path strategy on the ground station β (a)(b) , respectively assigned as a Task offloading rate during the outer cycle α (a) , Task offloading path strategy on satellite c (a) , Task offloading path strategy on the ground station β (a) ,in, α (a) 、β (a) 、c (a) The initial value of is randomly generated; (4) Offloading satellite communication tasks into the model α As the variable to be solved, b、c Assign values ​​to β (a)(b) , c (a)(b) , thereby optimizing the satellite communication task offloading model into a partial task offloading submodel, solving the partial task offloading submodel, and obtaining the firsta The satellite task offloading rate strategy during the b+1th outer cycle and the b+1th inner cycle α (a)(b+1) ; (5) Offloading satellite communication tasks into the model c As the variable to be solved, α Assigned to α (a)(b+1) , thereby optimizing the satellite communication task offloading model into a satellite offloading path selection sub-model; (6) Solve the satellite unloading path selection submodel and obtain a The unloading path strategy of tasks on the satellite during the b+1th outer cycle and the b+1th inner cycle c (a)(b+1) ; (7) Judgment α (a)(b+1) , c (a)(b+1) Whether the preset convergence condition is met, if so, execute step (8), otherwise, set b=b+1 and return to execute step (4); (8) Settings α (a+1) = α (a)(b+1) , c (a+1) = c (a)(b+1) ; (9) Offloading satellite communication tasks into the model β As the variable to be solved, a、c Assign values ​​to α (a+1) , c (a+1) , thereby optimizing the satellite communication task offloading model into a ground station offloading path selection sub-model; (10) Solve the ground station unloading path selection submodel and obtain a+ Task unloading path strategy on the ground station during one outer cycle β (a+1) ; (11) Judgment β (a+1) Is it equal to β (a) If not, then a=a +1,b=0, and return to step (3). If so, α (a+1) , β (a+1) , c (a+1) Output as the optimal value.

[0008] A computer program product comprises a computer program / instruction, wherein the computer program / instruction implements the above method when executed by a processor.

[0009] Compared with the prior art, the present invention has the following beneficial effects: the present invention takes into account the computational delay of satellite missions in a short period of time, regards the communication parameters in a short period of time as constants, and calculates the communication task volume and the interference on the path in advance, thereby optimizing multiple aspects including task offloading, path selection, and ground station task balance, thereby effectively reducing the total delay of task offloading, taking into account the actual situation of the satellite as much as possible, effectively overcoming the disadvantage of the limited computing power of LEO, greatly raising the upper limit of the amount of tasks accepted by a single LEO, and solving the problem of uneven distribution of tasks processed by ground stations, thus avoiding the problem of some ground stations being overloaded while some ground stations' computing resources being idle. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 It is a flowchart of a method for offloading near-Earth satellite communication tasks based on ground collaborative computing provided by an embodiment of the present invention; Figure 2 It is a satellite and ground station communication model diagram applicable to the present invention; Figure 3 It is a modeling diagram of the distance between the satellite and the ground station provided by the present invention; Figure 4 It is a flow chart of a method for solving a satellite unloading path selection sub-model provided by the present invention; Figure 5 It is a flow chart of a method for solving a ground station task offloading path selection sub-model provided by the present invention; Figure 6 This is the verification diagram of the Matlab simulation results. DETAILED DESCRIPTION

[0011] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0012] Embodiment 1.

[0013] The embodiment of the present invention provides a method for offloading near-earth satellite communication tasks based on ground collaborative computing, such as Figure 1 As shown, the following steps are included: (1) Establish a satellite communication task offloading model.

[0014] The communication model applicable to the present invention is as follows Figure 2As shown, in this model, multiple satellites communicate with a single ground station, multiple satellites communicate with each other, and multiple ground stations communicate with each other. When establishing a satellite communication task offloading model, the computing time of satellites and ground stations and the task transmission time are integrated, and the task volume processed by the ground station is balanced to avoid the problem of some ground stations being overloaded and some ground stations having idle computing resources due to changes in the satellite's geographical location. Therefore, the satellite communication task offloading model established by the present invention aims to minimize the total delay of task offloading, and the task offloading rate strategy of the satellite is used. α , Task offloading path strategy on satellite c , Task offloading path strategy on the ground station β is the variable to be solved, and the total delay of task offloading is specifically the sum of the calculation delay of the task on the satellite, the transmission delay of the task offloading from the satellite to the ground station, the transmission delay of the task between the ground stations and the calculation delay of the task on the ground station.

[0015] (1.1) The computational delay of the mission on the satellite.

[0016] The calculation delay of the task on the satellite is as follows: , In the formula, Indicates that the mission is on satellite s mn The computation delay on mn represents the mth satellite communicating with ground station n, Indicates satellite s mn The task offloading rate is Indicates satellite s mn The amount of tasks received, Indicates satellite s mn The computational amount of the received task, Indicates satellite s mn The complexity of the received task, Indicates satellite s mn The task arrival rate ， Indicates satellite s mn In the present invention, The value of is set to a variable that can take any value between 0 and 1. This setting is more flexible and reasonable than the binary value of 0 and 1, and can significantly reduce the total delay.

[0017] (1.2) Transmission delay of the task from satellite to ground station.

[0018] Since the scenario is satellite communication, the distance between the satellite and the ground station is large, so the time loss of the signal transmission path cannot be ignored. Introduce a task satellite unloading path indicator variable , The value is 0 or 1, when the mission is from the satellite s jn Unloaded to satellites mn hour, ,otherwise , s jn represents the jth satellite communicating with the nth ground station. Therefore, the transmission delay of the task offloading from the satellite to the ground station can be expressed as: , In the formula, Indicates that the mission is from satellite s mn The transmission delay to the ground station n is: Indicates satellite s mn The distance from the ground station n, represents the speed of light, Indicates satellite s jn The task offloading rate is Indicates satellite s jn The computational amount of the received task Indicates satellite s jn The task arrival rate ， Indicates satellite s mn The transmission rate of the task offloaded to the ground station, , Satellite s jn With satellite s mn The transmission rate and distance between represents the set of satellites communicating with ground station n, Indicates the transfer indicator variable. When a task is transmitted to the satellite s mn Then unloaded to ground station n ,otherwise , that is, when hour is 0 if yes, otherwise it is 1.

[0019] The derivation process is as follows: Satellite and ground station position modeling see Figure 3 ,in The satellite's altitude above the ground. It represents the angle between the line connecting the satellite ground station and the line connecting the satellite center of the earth. represents the radius of the Earth, represents the elevation angle between satellite ground stations, represents the angle between the satellite geocentric line and the ground station geocentric line. Figure 3 Using the law of sine we can deduce: , , Combining the above two formulas, we can deduce: , After finishing, we get: .

[0020] Using the law of cosines, the distance between the satellite and the ground station can be expressed as: .

[0021] The derivation process is as follows: Figure 3 , establish a spherical coordinate system to represent the positions of all satellites communicating with the same ground station: , represents the azimuth of the spherical coordinate system, then the position is Satellite s jn and the location is Satellite s mn The straight-line distance between them can be expressed as: , , The expressions are: ), ), Where B is the channel bandwidth, is the signal-to-noise ratio of satellite transmission, The present invention considers the satellite communication using the orthogonal frequency division multiple access (OFDMA) access technology, that is, does not consider the channel interference between satellites, then It can be expressed as: , In the formula, is the satellite-to-satellite transmission power, They represent the efficiency of satellite transmission antenna and satellite receiving antenna respectively, and They represent the gains of the transmitting antenna and the receiving antenna respectively. k represents the Boltzmann constant, T1 represents the temperature of the inter-satellite communication environment, represents the remaining noise interference between satellites, represents the free path loss between satellites. The expression of antenna gain in the formula is: , in, It indicates the diameter of the antenna. Indicates the wavelength of the transmitted signal.

[0022] The expression is: , Similar to the inter-satellite signal-to-noise ratio expression, It is expressed as: , In the formula, is the transmission power transmitted by the satellite to the ground station, represents the efficiency of the ground station receiving antenna, Indicates the gain of the ground station receiving antenna. T2 indicates the temperature of the satellite to ground station communication environment. represents the remaining noise interference between the satellite and the ground station, represents the free path loss between the satellite and the ground station. The specific expression is: .

[0023] Since there is rain attenuation caused by rain interference on the satellite ground station communication path, the following is considered: , In the expression, is the rain attenuation, where and is a coefficient determined by frequency and polarization mode, R is the precipitation rate, and L is the distance between the satellite and the ground station through the rain area, which can be expressed as: .

[0024] (1.3) The transmission delay of the satellite mission between ground stations.

[0025] Here, the communication network between ground stations is considered as a wide area network (WAN), and the task transmission process is also modeled as a queuing problem. As the transmission rate within the WAN, As the transmission bandwidth between ground stations, while considering optical fiber as the transmission medium, represents the distance between ground stations n and i. Then the transmission delay is considered to be: , In the formula, Indicates satellite mn The delay of the offloaded task from ground station n to ground station i is: the first part is the queuing delay, the second part is the transmission delay, and the third part is the transmission time on the path. N represents the set of ground stations. Indicates that the ground station l A collection of satellites for communications, Indicates that N excludes ground stations l The remaining ground stations Satellite s kl The task offloading rate, the computational amount of the received tasks, and the task arrival rate 。

[0026] Satellite s mn The transmission delay of the task between all ground stations is: , In the formula, Indicates satellite s mn The transmission delay of the task between ground stations, , Both represent the mission ground station unloading path indicator variables. s mn Unloaded to ground station n Unloaded to ground station again i hour, ,otherwise , Same reason.

[0027] (1.4) The computational delay of the satellite’s mission at the ground station.

[0028] Due to the uneven distribution of satellites, different ground stations need to process different tasks, which may cause some ground stations to have too much computing power while some ground stations have idle computing resources. Therefore, the present invention also considers the unloading of tasks between satellites to balance the computing power of ground stations. Using the M / M / 1 queuing model, we get satellite s mn The time it takes to calculate the task on ground station i for: , in, Indicates ground station i of computing resources, It is also the mission ground station unloading path indicator variable. Indicates satellite s klThe amount of tasks received, Satellite s kl The complexity of the received mission is s mn The computational delay of the task at the ground station for: .

[0029] (1.5) Total delay of offloading task.

[0030] Total delay of unloading tasks is the sum of the above delays. Therefore, the satellite communication task offloading model is established as follows: , , st , , , , , , , , , .

[0031] In the formula, For satellite s mn The maximum acceptable delay in task processing, For satellite s mn The time spent within the transmission range of the ground station, Indicates satellite s mn The total time it takes to offload the task to the ground station. α= { }, , .

[0032] (2) Set the number of outer loop iterations a =0, the number of inner loop iterations b=0.

[0033] (3) aThe satellite task offloading rate strategy during the bth outer cycle and the bth inner cycle α (a)(b) , Task offloading path strategy on satellite c (a)(b) , Task offloading path strategy on the ground station β (a)(b) , respectively assigned as a Task offloading rate strategy in the second outer loop α (a) , Task offloading path strategy on satellite c (a) , Task offloading path strategy on the ground station β (a) ,in, α (a) 、β (a) 、c (a) Initial value of α (0) 、β (0) 、c (0) Randomly generated.

[0034] (4) Offloading satellite communication tasks into the model α As the variable to be solved, b、c Assign values ​​to β (a)(b) , c (a)(b) , thereby optimizing the satellite communication task offloading model into a partial task offloading submodel, solving the partial task offloading submodel, and obtaining the first a The satellite task offloading rate strategy during the b+1th outer cycle and the b+1th inner cycle α (a)(b+1) .

[0035] Among them, the partial task offloading sub-model will b、c Assumed as a fixed value, assigned β (a)(b) , c (a)(b) ,Will As the only variable. The affected part, the partial task offloading sub-model is specifically: , st C1,C4,C5,C8,C9,C10, In the formula, Indicates that when adopting a strategy The total delay of Respectively The assignment of β (a)(b) , c (a)(b) middle The value of .

[0036] Since: 1) all items in the above model are linear functions or convex functions, the entire model can be regarded as a convex function; 2) the constraints are also linear or convex. Therefore, the convex optimization interior point method is used to solve it, and we can get The value of The values ​​of form a set, as the first a The satellite task offloading rate strategy during the b+1th outer cycle and the b+1th inner cycle α (a)(b+1) .

[0037] (5) Offloading satellite communication tasks into the model c As the variable to be solved, α Assigned to α (a)(b+1) , thereby optimizing the satellite communication task offloading model into a satellite offloading path selection sub-model.

[0038] The satellite unloading path selection sub-model is specifically: , , st C6,C7, In the formula, Indicates that when adopting a strategy The time delay, express The assignment of α (a)(b+1) middle The value of The value of Before calculating the delay, you need to first calculate the The situation is determined The value of .

[0039] (6) Solve the satellite unloading path selection submodel and obtain a The unloading path strategy of tasks on the satellite during the b+1th outer cycle and the b+1th inner cycle c (a)(b+1) .

[0040] like Figure 4 As shown, step (6) specifically includes: (6.1) Collect from ground station N Select any ground station n that has not been traversed yet; (6.2) Change the current strategy Updated to , where the current strategy represents the offloading path strategy of all satellites currently communicating with ground station n, is the offloading path strategy of all satellites communicating with ground station n at the end of the previous iteration, The initial value of is randomly generated; (6.3) Solve the following model to obtain the unloading path strategy of all satellites communicating with ground station n in the current iteration: , , st , , In the formula, Indicates that Smallest The value of Indicates that when adopting a strategy The total delay of the task received by ground station n is, represents the offloading path strategy of all satellites communicating with ground station n during the current iteration; (6.4) Calculation strategy The total delay of the task received by ground station n is , and adopt strategies The total delay of the task received by ground station n is ,in, express and The union of; (6.5) Determine whether , represents the time increment threshold. If it is satisfied, execute step (6.7), otherwise execute step (6.6); (6.6) = , and return to execute step (6.2); (6.7) As the optimal value of the offloading path strategy for all satellites communicating with ground station n ; (6.8) Determine the ground station set NWhether the traversal has been completed, if yes, execute step (6.9), if not, return to execute step (6.1); (6.9) As The optimal value of c (a)(b+1) .

[0041] (7) Judgment α (a)(b+1) , c (a)(b+1) Check whether the preset convergence condition is met. If so, execute step (8); otherwise, set b=b+1 and return to execute step (4).

[0042] The preset convergence condition is specifically: , In the formula, express The values ​​are The total delay of express The values ​​are The total delay of Indicates the latency threshold.

[0043] (8) Settings α (a+1) = α (a)(b+1) , c (a+1) = c (a)(b+1) .

[0044] (9) Offloading satellite communication tasks into the model β As the variable to be solved, a、c Assign values ​​to α (a+1) , c (a+1) , thereby optimizing the satellite communication task offloading model into a ground station offloading path selection sub-model.

[0045] The ground station unloading path selection sub-model is specifically: , , In the formula, represents the ground station unloading path selection sub-model, specifically a game model, V represents the set of ground stations participating in the game, Satellites mn The utility function of the offloaded task, 、 Respectively 、 The assignment of α (a+1) middle 、 The value of .

[0046] By executing the proposed latent game, multiple ground stations collaborate to determine where each vehicle's offloading task is processed. Each ground station is responsible for selecting the best strategy improvement for each satellite offloading task it covers.

[0047] (10) Solve the ground station unloading path selection submodel and obtain a+ Task unloading path strategy on the ground station during one outer cycle β (a+1) .

[0048] In this step, a multi-round game method is used to solve the ground station unloading path selection sub-model, such as Figure 5 As shown in Figure 1, each round of the game specifically includes: (10.1) Change the current strategy Update to policy , where the current strategy The initial value of is randomly generated. is the unloading path strategy at the end of the previous iteration; (10.2) For each ground station n, solve the following model to obtain the offloading path strategy for each satellite communicating with ground station n: , , In the formula, Indicates satellites mn Uninstall path strategy, Indicates that Smallest The value of (10.3) Add the unloading path strategies of all satellites to the total strategy set Ω and solve the following model: , , In the formula, represents the total utility, Indicates that Smallest The value of (10.4) Determine whether ,in, Indicates that join in Replace the original The total utility after Indicates that in the current strategy The total utility under ; if yes, execute step (10.5), otherwise execute step (10.6); (10.5) Strategy Update to the solution obtained in step (10.3) , and Add to Ω and return to step (10.3); (10.6) Judgment strategy Is it equal to the current strategy? If not, return to step (10.1). If equal, the game has reached Nash equilibrium. As The final strategy is assigned to β (a+1) .

[0049] (11) Judgment β (a+1) Is it equal to β (a) , if not, then a=a +1,b=0, and return to step (3). If so, α (a+1) , β (a+1) , c (a+1) As the optimal value α* , β* , c* Output.

[0050] The invention is verified by MATLAB simulation below.

[0051] In order to verify the feasibility and optimize the performance of the present invention, Figure 6 The simulation results of MATLAB are shown. The blue line is the delay caused by using the method of the present invention; the red line is the delay of offloading the task from the local satellite to the ground station for calculation without considering inter-satellite communication. By observing the simulation results, it is found that when processing different satellite task amounts, the present invention significantly reduces the delay of task processing compared to the solution of offloading all tasks to the ground station, which proves that the present invention can bring lower satellite task processing delay.

[0052] The embodiment of the present invention aims to provide a method for reducing the delay in near-Earth satellite communications by task offloading and resource allocation based on game theory. In view of the problems of excessive satellite tasks, the influence of rain on satellite transmission paths, and high loads on some ground stations while idle computing resources on some ground stations, a two-stage iterative algorithm is designed to significantly reduce the delay in task processing, balance the task load of each ground station, and optimize the total delay in task processing of multiple satellites and multiple ground stations. The present invention is compared with other solutions through simulation to verify the better optimization performance of the present invention. At the same time, the present invention is expected to be combined with the currently studied scheme for optimizing the task offloading performance within the satellite cycle time, so as to achieve greater optimization of satellite communication delay, task download rate and other aspects.

[0053] Embodiment 2.

[0054] The embodiment of the present invention also provides a computer product, such as an app on a mobile phone or a tablet, an installation program on a computer, etc., which includes a computer program / instruction, and the computer program / instruction is executed by a processor to implement the method described in Example 1. The code of the computer executable program for performing the operation of the present invention can be written in one or more programming languages ​​or a combination thereof, and the programming language includes an object-oriented programming language, such as Java, Smalltalk, C++, and also includes a conventional procedural programming language, such as "C" language or similar programming language. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0055] It should be understood that the above embodiments and descriptions only describe the principles, main features and advantages of the present invention. Without departing from the spirit and scope of the present invention, the present invention may be subject to various changes and improvements, and these changes and improvements all fall within the scope of protection of the present invention.

Claims

1. A method for offloading near-Earth satellite communication tasks based on ground collaborative computing, characterized in that: The steps include: (1) Establishing a satellite communication task offloading model, wherein the satellite communication task offloading model aims to minimize the total delay of task offloading and takes the task offloading rate strategy of the satellite as the α , Task offloading path strategy on satellite γ , Task offloading path strategy on the ground station β is the variable to be solved, and the total delay of task offloading is specifically the sum of the calculation delay of the task on the satellite, the transmission delay of the task offloading from the satellite to the ground station, the transmission delay of the task between the ground stations and the calculation delay of the task on the ground station; (2) Set the number of outer loop iterations a =0, the number of inner loop iterations b=0; (3) a The satellite's mission offloading rate strategy during the bth outer cycle and the bth inner cycle α (a)(b) , Task offloading path strategy on satellite γ (a)(b) , Task offloading path strategy on the ground station β (a)(b) , respectively assigned as a Task offloading rate strategy in the second outer loop α (a) , Task offloading path strategy on satellite γ (a) , Task offloading path strategy on the ground station β (a) ,in, α (a) 、β (a) , γ (a) The initial value of is randomly generated; (4) Offloading satellite communication tasks into the model α As the variable to be solved, β, γ Assign values ​​to β (a)(b) , γ (a)(b) , thereby optimizing the satellite communication task offloading model into a partial task offloading submodel, solving the partial task offloading submodel, and obtaining the first a The satellite task offloading rate strategy during the b+1th outer cycle and the b+1th inner cycle α (a)(b+1) ; (5) Offloading satellite communication tasks into the model γ As the variable to be solved, α Assigned to α (a)(b+1) , thereby optimizing the satellite communication task offloading model into a satellite offloading path selection sub-model; (6) Solve the satellite unloading path selection submodel and obtain a The unloading path strategy of tasks on the satellite during the b+1th outer cycle and the b+1th inner cycle γ (a)(b+1) ; (7) Judgment α (a)(b+1) , γ (a)(b+1) Whether the preset convergence condition is met, if so, execute step (8), otherwise, set b=b+1 and return to execute step (4); (8) Settings α (a+1) = α (a)(b+1) , γ (a+1) = γ (a)(b+1) ; (9) Offloading satellite communication tasks into the model β As the variable to be solved, α, γ Assign values ​​to α (a+1) , γ (a+1) , thereby optimizing the satellite communication task offloading model into a ground station offloading path selection sub-model; (10) Solve the ground station unloading path selection submodel and obtain a+ Task unloading path strategy on the ground station during one outer cycle β (a+1) ; (11) Judgment β (a+1) Is it equal to β (a) , if not, then a=a +1,b=0, and return to step (3). If so, α (a+1) , β (a+1) , γ (a+1) As α, β, γ The optimal value output.

2. The method for offloading near-Earth satellite communication tasks based on ground collaborative computing according to claim 1 is characterized in that: The satellite communication task offloading model is specifically as follows: ， ， ， ， ， ， s.t. , ， ， ， ， ， ， ， ， ， In the formula, t represents the total delay of task offloading, 、 Respectively represent the ground station n, l , i A collection of satellites for communications, N represents the set of ground stations, Represents ground station n and satellite s mn The unloading delay of the task between mn represents the mth satellite communicating with ground station n, Indicates that the mission is on satellite s mn The computational delay on Indicates that the mission is from the satellite s mn The transmission delay to offload to ground station n, Indicates satellite s mn The transmission delay of the task between ground stations, Indicates satellite s mn The computational delay of the task at the ground station; , Satellite s mn 、s jn 、s kl The task offloading rate is s jn represents the jth satellite communicating with the nth ground station, s kl Indicates l The kth satellite that the ground station communicates with, Indicates satellite s mn The amount of tasks received, Indicates satellite s kl The amount of tasks received, Satellite s mn 、s jn 、s kl The computational amount of the received task, Satellite s mn 、s kl The complexity of the received task, Satellite s mn 、s jn 、s kl The task arrival rate ， Indicates satellite s mn of computing resources, Indicates satellite s mn The distance from the ground station n, represents the speed of light, Indicates the mission satellite unloading path indicator variable with a value of 0 or 1. s jn Unloaded to satellite s mn hour ,otherwise , Indicates the transfer indicator variable. When a task is transmitted to the satellite s mn Then unloaded to ground station n ,otherwise , Indicates satellite s mn The transmission rate of the task offloaded to the ground station, , Satellite s jn With satellite s mn The transmission rate and distance between , Both represent the mission ground station unloading path indicator variables. s mn Unloaded to ground station n Unloaded to ground station again i hour ,otherwise , represents the transmission rate between ground stations, represents the bandwidth of transmission between ground stations, represents ground station n and ground station i The distance between Indicates ground station i of computing resources, For satellite s mn The maximum acceptable delay in task processing, For satellite s mn The time spent within the transmission range of the ground station, Indicates satellite s mn The total time it takes to offload the task to the ground station. α= { }， ， 。 3. The method for offloading near-Earth satellite communication tasks based on ground collaborative computing according to claim 2 is characterized in that: The partial task offloading sub-model described in step (4) is specifically: ， st C1,C4,C5,C8,C9,C10, In the formula, Indicates that when adopting a strategy The total delay of Respectively The assignment of β (a)(b) middle The value of γ (a)(b) middle The value of .

4. The method for offloading near-Earth satellite communication tasks based on ground collaborative computing according to claim 2, characterized in that: The partial task offloading sub-model described in step (4) is solved using the convex optimization interior point method.

5. The method for offloading near-Earth satellite communication tasks based on ground collaborative computing according to claim 2, characterized in that: The satellite unloading path selection sub-model in step (5) is specifically: ， st C6,C7, In the formula, Indicates that when adopting a strategy The time delay, express The assignment of α (a)(b+1) middle The value of .

6. The method for offloading near-Earth satellite communication tasks based on ground collaborative computing according to claim 5, characterized in that: Step (6) specifically includes: (6.1) Collect from ground station N Select any ground station n that has not been traversed yet; (6.2) Change the current strategy Updated to , where the current strategy represents the offloading path strategy of all satellites currently communicating with ground station n, is the offloading path strategy of all satellites communicating with ground station n at the end of the previous iteration, The initial value of is randomly generated; (6.3) Solve the following model to obtain the unloading path strategy of all satellites communicating with ground station n in the current iteration: ， ， s.t. , ， In the formula, Indicates that Smallest The value of Indicates that when adopting a strategy The total delay of the task received by ground station n is, represents the offloading path strategy of all satellites communicating with ground station n during the current iteration; (6.4) Calculation strategy The total delay of the task received by ground station n is , and adopt strategies The total delay of the task received by ground station n is ,in, express and The union of; (6.5) Determine whether , represents the time increment threshold. If it is satisfied, execute step (6.7), otherwise execute step (6.6); (6.6) = , and return to execute step (6.2); (6.7) As the optimal value of the offloading path strategy for all satellites communicating with ground station n ; (6.8) Determine the ground station set N Whether the traversal has been completed, if yes, execute step (6.9), if not, return to execute step (6.1); (6.9) As The optimal value of γ (a)(b+1) .

7. The method for offloading near-Earth satellite communication tasks based on ground collaborative computing according to claim 2, characterized in that: The ground station unloading path selection sub-model in step (9) is specifically: ， ， In the formula, represents the ground station unloading path selection sub-model, specifically a game model, V represents the set of ground stations participating in the game, Satellites mn The utility function of the offloaded task, 、 Respectively 、 The assignment of α (a+1) middle 、 The value of .

8. The method for offloading near-Earth satellite communication tasks based on ground collaborative computing according to claim 7 is characterized in that: In step (10), a multi-round game method is used to solve the ground station unloading path selection sub-model, and each round of the game specifically includes: (10.1) Change the current strategy Update to policy , where the current strategy The initial value of is randomly generated. is the unloading path strategy at the end of the previous iteration; (10.2) For each ground station n, solve the following model to obtain the offloading path strategy for each satellite communicating with ground station n: , ， In the formula, Indicates satellites mn Uninstall path strategy, Indicates that Smallest The value of (10.3) Add the unloading path strategies of all satellites to the total strategy set Ω and solve the following model: ， ， In the formula, represents the total utility, Indicates that Smallest The value of (10.4) Determine whether ,in, Indicates that join in Replace the original The total utility after Indicates that in the current strategy The total utility under ; if yes, execute step (10.5), otherwise execute step (10.6); (10.5) Strategy Update to the solution obtained in step (10.3) , and Add to Ω and return to step (10.3); (10.6) Judgment strategy Is it equal to the current strategy? If not, return to step (10.1). If equal, the game has reached Nash equilibrium. As The final strategy is assigned to β (a+1) .

9. The method for offloading near-Earth satellite communication tasks based on ground collaborative computing according to claim 2, characterized in that: The preset convergence condition in step (7) is specifically: ， In the formula, express The values ​​are The total delay of express The values ​​are The total delay of Indicates the latency threshold.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the method according to any one of claims 1 to 9 is implemented.

Images