A computational method for migrating blocked tasks in low-orbit satellite computing networks

Abstract

The present invention belongs to the field of electronic information, and discloses a calculation method for the migration of blocked tasks in a low-orbit satellite computing network, which counts the system parameters of the low-orbit satellite computing network and calculates the parameters related to task processing and transmission; then, the maximum task blocking probability and the maximum new task arrival rate of the low-orbit satellite are calculated according to the parameters, and the value range of the migration parameter under the corresponding situation is determined, so as to find the optimal value of the migration parameter by optimizing the average task completion time; according to the optimal migration parameter, the tasks blocked by the low-orbit satellite are assigned between adjacent satellites and ground users, and the system parameters are continued to be counted, and whether to recalculate the migration parameter is determined by judging the size of the parameter exceeding the error range. The present invention performs migration scheduling for each blocked task between satellites and between satellites and users, so as to effectively utilize the computing resources of the low-orbit satellite network.

Description

Technical Field

[0001] The invention belongs to the field of electronic information, and in particular relates to a calculation method for migrating blocked tasks in a low-orbit satellite computing network. Background Art

[0002] In the satellite-ground integrated network, low-orbit satellites can provide ground users with a variety of services such as communications, edge computing and remote sensing. However, remote sensing and other services require low-orbit satellites to provide high-definition detection images with large amounts of data. Low-orbit satellites usually need to spend a long time transmitting and transmitting power to transmit these unprocessed data to ground users. Therefore, in order to reduce transmission time and energy consumption, low-orbit satellites can use onboard computing resources to process unprocessed data and only send the processed results to ground users.

[0003] Due to the size, weight and power limitations of low-orbit satellites, the onboard computing resources of a single satellite are usually limited. As the number of users increases, low-orbit satellites require more onboard computing resources to handle these tasks. To solve this problem, low-orbit satellites can use inter-satellite links to form a network, thereby sharing computing resources and improving computing efficiency through collaborative computing. Therefore, the task migration method can effectively schedule computing tasks between low-orbit satellites and between low-orbit satellites and ground users, thereby improving the utilization efficiency of computing resources.

[0004] Although the current task migration method can achieve the optimal scheduling of computing resources, it does not take into account the following issues. First, task migration usually requires obtaining the status information of adjacent low-orbit satellites. However, the distance between satellites is far, and the acquisition of status information requires increased costs and the acquired information has a lag. Second, in practical applications, due to the changes in the number of users, power management, and computing resource sharing faced by some computing nodes in low-orbit satellites, the randomness of task arrival and task processing time changes dynamically. Finally, task migration will lead to an increase in the number of computing tasks in the low-orbit satellite system. When the number of computing tasks exceeds the system's carrying capacity, it will have a negative impact on the computing stability of the low-orbit satellite network. Summary of the invention

[0005] To solve the above technical problems, the present invention provides a calculation method for migrating blocked tasks in a low-orbit satellite computing network, which is applied to the low-orbit satellite computing network to schedule the migration of blocked tasks between low-orbit satellites and between low-orbit satellites and users. The method does not need to obtain the status information of adjacent low-orbit satellites in real time, thus avoiding the impact of task migration on the computing stability of the low-orbit satellite network, and is adaptable to scenarios with dynamic changes in task arrival and task processing time.

[0006] The method for calculating the migration of blocked tasks in a low-orbit satellite computing network according to the present invention comprises the following steps:

[0007] Step 1: Count the fixed time in the low-orbit satellite calculation network Number of new missions to reach LEO satellites within , and calculate the system parameters related to the arrival rate and transmission time of new tasks;

[0008] Step 2: Calculate the maximum mission blocking probability of each low-orbit satellite and the maximum new mission arrival rate of the low-orbit satellite using the system parameters obtained in step 1;

[0009] Step 3: compare the new task arrival rate with the maximum new task arrival rate, and calculate the value range of the migration parameter in the corresponding case, where the migration parameter refers to the probability that the low-orbit satellite will send the blocked task to any adjacent earth-orbit satellite;

[0010] Step 4: Calculate the optimal value of the migration parameter according to the value range of the migration parameter so as to minimize the average task completion time;

[0011] Step 5: Allocate tasks blocked by low-orbit satellites among adjacent low-orbit satellites and ground users based on the optimal value of the migration parameter;

[0012] Step 6: Continue to count the new task arrival rate and system parameters in step 1 to determine whether the new task arrival rate and system parameters exceed the preset error range; if so, repeat steps 2 to 5 based on the new system parameters obtained.

[0013] Furthermore, in step 1, the system parameters include the average time for transmitting unprocessed data to ground users , the average time to transmit the processed data results to ground users , the average time to transmit unprocessed data to nearby low-orbit satellites and the average time for low-orbit satellites to process tasks ;

[0014] The new task arrival rate ; A new mission refers to the data acquisition request from ground users that arrives at any low-orbit satellite for the first time, where the data required by the user refers to the data generated in the low-orbit satellite; Low-orbit satellites complete their missions in the following two ways: using onboard computing resources to process data and sending the processed data results back to ground users; not performing data processing and directly sending unprocessed data back to ground users.

[0015] Furthermore, in step 2, task blocking refers to the low-orbit satellite refusing or failing to provide computing services for the task; the task blocking probability P of the low-orbit satellite is defined as: the probability of any task being blocked when the low-orbit satellite receives a newly arrived task and a blocked task sent by other low-orbit satellites;

[0016] Maximum task blocking probability It is defined as: the maximum value of the mission blocking probability P of the low-orbit satellite, that is, when the total mission arrival rate Equal to the total computing speed of all computing nodes in the low-orbit satellite The probability that any mission is blocked by a low-orbit satellite is: , Σ is the summation symbol, and the variable i in the subsequent formula ranges from 0 to The sum of all integer values ​​between;

[0017] Among them, the total mission includes new missions and blocked missions sent by other low-orbit satellites. is the number of computing nodes in the LEO satellite;

[0018] Maximum new mission arrival rate for LEO satellites The calculation method is: the total task arrival rate satisfies Under the conditions, using Calculate the ratio of new tasks to the total number of tasks, that is, .

[0019] Furthermore, step 3 is specifically as follows:

[0020] if , then the migration parameters The value range is ,in ,Right now Take any value between 0 and 1. Represents relationships between elements and sets;

[0021] if , then the migration parameters The value range of is pre-set in the range of 0 to 1 according to the actual situation or calculated according to the following formula: .

[0022] Furthermore, step 4 is specifically as follows:

[0023] The average task completion time T is calculated as follows: based on the probability of the task being served, blocked, migrated to an adjacent low-orbit satellite, and sent to the user when the task arrives at the low-orbit satellite, as well as the corresponding time, the average time taken for the task to arrive at the low-orbit satellite and for the ground user to receive the data processing results or unprocessed data is calculated; the average task completion time , the formula is:

[0024] ,

[0025] At this time, the total task arrival rate , the formula is: ;

[0026] Optimal values ​​of migration parameters The calculation method is: based on the average task completion time , the search makes Migration parameters that reach their minimum The value of is: ,

[0027] Here, argmin represents the parameter value at which a function achieves the minimum value in its domain.

[0028] Furthermore, in step 5, when any task is blocked by a low-orbit satellite, the low-orbit satellite The blocked task is randomly sent to any adjacent low-orbit satellite with a probability of The probability of giving up processing the blocked task is 0.01, and the unprocessed data is directly sent to the ground user to which the task belongs.

[0029] Furthermore, in step 6, the system parameters related to the arrival rate and transmission time of the new task are re-statisticed at a certain interval; if the newly obtained data exceeds the set error range, the new data is used to repeat steps 2 to 5 to update the optimal value of the migration parameter. , and according to the updated Allocate the blocked task.

[0030] The beneficial effects described in the present invention are as follows: the method described in the present invention migrates blocked tasks in a low-orbit satellite computing network, and performs system parameter statistics according to the statistical method of step 1, without the need to obtain the status information of adjacent low-orbit satellites in real time, thereby reducing the cost required to obtain the status information; the method of the present invention reduces the potential negative impact of blocked task migration on the computing stability of the overall low-orbit satellite network by limiting the value range of migration parameters; the computing resource utilization efficiency of the low-orbit satellite network is improved by performing migration scheduling for each blocked task between low-orbit satellites and between low-orbit satellites and users; the setting of system parameters and parameter error ranges is performed regularly, and the migration parameters are adjusted to dynamically adapt to the scenario of random changes in task arrival and task processing time. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a schematic diagram of the application of the method of the present invention;

[0032] Figure 2 is a flow chart of the method of the present invention;

[0033] Figure 3 It is a schematic diagram of the migration parameter calculation and blocked task migration process;

[0034] Figure 4 It is a schematic diagram of the simulation results of the task completion time changing with the arrival rate of new tasks;

[0035] Figure 5 It is the simulation result of processing other data type scenarios;

[0036] Figure 6 This is the simulation result under the scenario of high transmission cost between low-orbit satellites. DETAILED DESCRIPTION

[0037] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments in conjunction with the accompanying drawings.

[0038] like Figure 1 As shown, a calculation method for migrating blocked tasks in a low-orbit satellite computing network described in the present invention is applied to a low-orbit satellite computing network. When a task reaching a satellite is blocked, the blocked task is randomly sent to an adjacent low-orbit satellite through an algorithm, or the unprocessed data is directly sent to ground users.

[0039] like Figure 2 and Figure 3 As shown, a calculation method for migrating blocked tasks in a low-orbit satellite computing network according to the present invention comprises the following steps:

[0040] Step 1: System parameter statistics: Methods for counting relevant parameters in the low-orbit satellite computing network, for example, statistics of fixed time Number of new missions to reach LEO satellites within , then the new task arrival rate Similarly, the average time for transmitting unprocessed data to ground users is calculated , the average time to transmit the processed data results to ground users , the average time to transmit unprocessed data to nearby low-orbit satellites and the average time for low-orbit satellites to process tasks .

[0041] Step 2: Calculation of the maximum new task arrival rate: Consider the number of computing nodes in the low-orbit satellite is , and the low-orbit satellite does not have enough memory to queue the arriving tasks, when the total task arrival rate Equal to the total computing speed of all computing nodes in the low-orbit satellite The probability that any mission is blocked by a low-orbit satellite , the formula is:

[0042] (1),

[0043] Therefore, the maximum new mission arrival rate of LEO satellites , the formula is:

[0044] (2).

[0045] Step 3: Calculate the range of migration parameter values: If , then the migration parameters The value range is ,in , It can take any value in the range of 0 to 1;

[0046] if , then the migration parameters The value range of is:

[0047] (3).

[0048] Step 4: Optimal migration parameter value: average task completion time , the formula is:

[0049] (4),

[0050] Among them, P is the mission blocking probability of the low-orbit satellite, which is defined as the probability that any mission is blocked by a low-orbit satellite when the low-orbit satellite receives a blocked mission sent by other low-orbit satellites. The formula is:

[0051] (5),

[0052] in, is the total task arrival rate, and its formula is:

[0053] (6);

[0054] Optimal values ​​of migration parameters , the formula is:

[0055] (7),

[0056] Among them, argmin represents the parameter value of a function that takes the minimum value in its domain. Represents the relationship between elements and collections. The calculation of can be done by the following approximate algorithm: within the value range R, from Start by increasing the value by 0.01 each time. Substitute it into formula (4) for calculation. After calculating all the results, the value of the minimum average task completion time T corresponds to The value of .

[0057] Step 5: Migrate blocked tasks: When any task is blocked by a low-orbit satellite, the low-orbit satellite will The mission will be randomly sent to any adjacent low-orbit satellite. The probability of giving up processing the blocked task is 0.01, and the unprocessed data is directly sent to the ground user to which the task belongs.

[0058] Step 6: System parameter statistics: Use the method in step 1 to re-count each parameter. Set an error range for each parameter, such as the arrival rate of new tasks. The error range is ; If any newly obtained system parameter exceeds the set error range, such as the re-counted new task arrival rate or , then use the new system parameters to update the optimal values ​​of the migration parameters according to steps 2 to 5 , and according to the updated Allocate the blocked task.

[0059] The method described in the present invention is applied to a low-orbit satellite computing network. First, based on the simulation parameter settings shown in Table 1, the task completion time using the method described in the present invention and not using the method described in the present invention is compared. Second, by changing the simulation parameters, the effectiveness of the algorithm of the present invention in different environments is verified.

[0060] Table 1 Simulation example parameter settings

[0061]

[0062] according to Figure 4 As shown in FIG. 1 , the task completion time increases with the increase of the new task arrival rate. When the new task arrival rate is low, the method of the present invention can reduce the task completion time by about 30% at most. Even when the new task completion time is large, the method of the present invention can still significantly improve the computing performance of the low-orbit satellite computing system.

[0063] (1) Changes in the time distribution of low-orbit satellite processing tasks:

[0064] Simulate the scenario when the low-orbit satellite processes other data. At this time, the task processing time changes, changing the average time of low-orbit satellite processing tasks in Table 1 The Pareto distribution is used to sample the time of low-orbit satellite processing tasks, where the scale parameter of the Pareto distribution is 0.3, the shape parameter is 1.5, and the time mean is 0.9s. Other simulation parameters remain unchanged.

[0065] according to Figure 5As shown in Figure 2, when the arrival rate of new tasks is low, the use of this method can reduce the task completion time by up to about 20%.

[0066] (2) Changes in the average time to transmit unprocessed data to nearby satellites:

[0067] Simulate the application scenario where the transmission cost between low-orbit satellites is high, and increase the average time of transmitting unprocessed data to nearby satellites in Table 1 , so that At this time, using the method described in the present invention to migrate the blocked mission to a nearby satellite has a large cost.

[0068] according to Figure 6 As shown, in this environment, although the method of the present invention has limited improvement on the performance of the low-orbit satellite computing system, when the arrival rate of new tasks is low, the method of the present invention can still reduce the task completion time.

[0069] The experimental results illustrate the feasibility and practicality of the method of the present invention. Through simulation results, it is verified that under different environmental parameters, when the method of the present invention is used, the task calculation time of the low-orbit satellite computing network can be effectively reduced. Especially when the task arrival rate is small, the method of the present invention can further improve the computing performance of the low-orbit satellite network.

[0070] The above description is only a preferred embodiment of the present invention and is not intended to be a further limitation of the present invention. All equivalent changes made using the contents of the present specification and drawings are within the protection scope of the present invention.

Claims

1. A computational method for migrating blocked tasks in a low-orbit satellite computing network, characterized in that: The following steps are involved: Step 1: Count the fixed time in the low-orbit satellite calculation network Number of new missions to reach LEO satellites within , and calculate the system parameters related to the arrival rate and transmission time of new tasks; The system parameters include the average time to transmit unprocessed data to ground users. , the average time to transmit the processed data results to ground users , the average time to transmit unprocessed data to nearby low-orbit satellites and the average time for low-orbit satellites to process tasks ; Step 2: Calculate the maximum mission blocking probability of each low-orbit satellite and the maximum new mission arrival rate of the low-orbit satellite using the system parameters obtained in step 1; specifically: Task blocking refers to the refusal or inability of a low-orbit satellite to provide computing services for the task. The task blocking probability P of a low-orbit satellite is defined as: the probability of any task being blocked when the low-orbit satellite receives a newly arrived task and a blocked task sent by other low-orbit satellites; Maximum task blocking probability It is defined as: the maximum value of the mission blocking probability P of the low-orbit satellite, that is, when the total mission arrival rate Equal to the total computing speed of all computing nodes in the low-orbit satellite The probability that any mission is blocked by a low-orbit satellite is: , Σ is the summation symbol, and the variable i in the subsequent formula ranges from 0 to The sum of all integer values ​​between; Among them, the total mission includes new missions and blocked missions sent by other low-orbit satellites. is the number of computing nodes in the LEO satellite; Maximum new mission arrival rate for LEO satellites The calculation method is: the total task arrival rate satisfies Under the conditions, using Calculate the ratio of new tasks to the total number of tasks, that is, ; Step 3: compare the new task arrival rate with the maximum new task arrival rate, and calculate the value range of the migration parameter in the corresponding case, where the migration parameter refers to the probability that the low-orbit satellite will send the blocked task to any adjacent earth-orbit satellite; specifically: if , then the migration parameters The value range is ,in ,Right now Take any value between 0 and 1. Represents relationships between elements and sets; if , then the migration parameters The value range of is pre-set in the range of 0 to 1 according to the actual situation or calculated according to the following formula: , ; Step 4: Calculate the optimal value of the migration parameter according to the value range of the migration parameter so as to minimize the average task completion time; Step 5: Allocate tasks blocked by low-orbit satellites among adjacent low-orbit satellites and ground users based on the optimal value of the migration parameter; Step 6: Continue to count the new task arrival rate and system parameters in step 1 to determine whether the new task arrival rate and system parameters exceed the preset error range; if so, repeat steps 2 to 5 based on the new system parameters obtained.

2. The method for calculating the migration of blocked tasks in a low-orbit satellite computing network according to claim 1, characterized in that: In step 1, The new task arrival rate ; A new mission refers to the data acquisition request from ground users that arrives at any low-orbit satellite for the first time, where the data required by the user refers to the data generated in the low-orbit satellite; Low-orbit satellites complete their missions in the following two ways: using onboard computing resources to process data and sending the processed data results back to ground users; not performing data processing and directly sending unprocessed data back to ground users.

3. The method for calculating the migration of blocked tasks in a low-orbit satellite computing network according to claim 2, characterized in that: Step 4 is as follows: Average task completion time The calculation formula is: ， At this time, the total task arrival rate , the formula is: ； Optimal values ​​of migration parameters The calculation method is: ， Here, argmin represents the parameter value at which a function achieves the minimum value in its domain.

4. The method for calculating blocked task migration in a low-orbit satellite computing network according to claim 3, characterized in that: In step 5, when any task is blocked by a low-orbit satellite, the low-orbit satellite The blocked task is randomly sent to any adjacent low-orbit satellite with a probability of The probability of giving up processing the blocked task is 0.01, and the unprocessed data is directly sent to the ground user to which the task belongs.

5. The method for calculating the migration of blocked tasks in a low-orbit satellite computing network according to claim 4, characterized in that: In step 6, the system parameters related to the arrival rate and transmission time of the new task are re-stated at a certain interval; if the newly obtained data exceeds the set error range, the new data is used to repeat steps 2 to 5 to update the optimal value of the migration parameter. , and according to the updated Allocate the blocked task.

Images