Task offloading method and device for low earth orbit satellite and ground edge computing network

Abstract

The application provides a low earth orbit satellite and ground edge computing network task offloading method and device, belonging to the field of wireless communication, which comprises the following steps: determining the current state of device offloading allocation, calculating the total system cost of the current state; for each device, cyclically traversing the decision result of processing the to-be-processed task locally or offloading to the base station or satellite for processing, and determining the decision result with the minimum total system cost in all decision results, and selecting the candidate decision result from the decision results of all devices; if the total system cost of the candidate decision result is less than the total system cost of the current state, updating the current state; repeating the above updating process until the total system cost of the candidate decision of each device is greater than the current state. The game theory decision method provided by the method has excellent performance compared with the commonly used benchmark method, can greatly reduce the system energy consumption, and improves the task offloading efficiency.

Description

Technical Field

[0001] This invention relates to the field of wireless communication, and in particular to a method and apparatus for offloading tasks between a low Earth orbit satellite and a ground-based edge computing network. Background Technology

[0002] With the unprecedented development of emerging applications such as the Internet of Things, virtual reality, and 4K video transmission, traditional terrestrial networks are struggling to meet the demands of integrated global air-space-ground communication because they cannot fully cover complex terrains such as remote mountainous areas and oceans. Furthermore, terrestrial network infrastructure is vulnerable to damage from natural disasters such as earthquakes and hurricanes, which can disrupt device communication.

[0003] In recent years, with the development of space communication networks, satellite technology has made significant progress in commercial, civilian, and military services, enabling the miniaturization of satellites, especially low-Earth orbit (LEO) satellites. Several major LEO satellite projects have been launched, such as OneWeb, SpaceX Starlink, and O3B, and recent news indicates that SpaceX will collaborate with Azure to provide global communication services for these devices. Therefore, it can be concluded that LEO satellite communication networks have unparalleled advantages over terrestrial mobile communication systems, achieving seamless global coverage and becoming an indispensable part of daily life.

[0004] On the other hand, the increasing prevalence of mobile devices (such as smartphones and tablets) has spurred the development of many new computationally intensive applications, such as speech recognition, gaming, multimedia encoding / decoding, and intelligent transportation. Therefore, low-Earth orbit (LEO) satellite networks not only need to provide channel access for these devices but also support various computing services. Generally, due to the limited computing and storage resources of ground-based equipment, computing tasks can be offloaded to ground base stations with abundant computing resources. Furthermore, computing tasks generated by equipment in remote areas and at sea without the support of ground network communication infrastructure can only be processed through servers deployed on LEO satellite networks. However, due to the altitude limitations of LEO satellites, the transmission latency of devices in LEO satellite networks increases accordingly, making it difficult to meet the real-time requirements of ground-based equipment. Summary of the Invention

[0005] To address the problems existing in the prior art, the present invention provides a method and apparatus for offloading tasks between low Earth orbit satellites and ground edge computing networks.

[0006] This invention provides a task offloading method between low Earth orbit satellites and ground-based edge computing networks, comprising: determining an initial decision result for all devices to process tasks locally, offload them to a base station, or offload them to a low Earth orbit satellite, as the current state; calculating the total cost of the system comprised of all devices in the current state; for each device, iterating through all decision results for processing tasks locally, offloading them to a base station, or offloading them to a low Earth orbit satellite; determining the decision result that minimizes the total system cost among all decision results for each device, and selecting a candidate decision result from all decision results for all devices; if the total system cost corresponding to the candidate decision result is less than the total system cost of the current state, updating the current state according to the candidate decision result; repeating the process of iterating through decision results for each device until updating the current state according to the candidate decision result, until the total system cost corresponding to all decision results for each device is not less than the total system cost of the current state; wherein, the total system cost is determined based on the cost of all devices; the local computing cost of each device is determined based on processor resource requirements; the cost for each device to offload tasks to a base station or a low Earth orbit satellite is determined based on the corresponding transmission energy consumption.

[0007] According to the task offloading method for low Earth orbit satellites and ground edge computing networks provided by the present invention, the intra-cell interference caused by other devices on the same channel to any device selecting low Earth orbit satellites for task offloading is less than a preset threshold; the total delay of any device selecting low Earth orbit satellites for task offloading is less than a preset threshold; the intra-cell power interference of any device to all devices on the same base station and its channel is less than a preset threshold; and the total computing power provided by each base station to devices within its coverage area is less than the maximum computing power of each base station.

[0008] A task offloading method for low-Earth orbit satellites and ground edge computing networks provided by the present invention further includes calculating the cost of each device according to the following formula:

[0009]

[0010]

[0011] Among them, h i For device u i The resulting decision, h -i Indicates excluding u i Decisions for all other equipment; C i (h i h -i ) indicates device u i Uninstallation decision h i The resulting uninstallation costs, These represent the costs of offloading the mission via low Earth orbit satellites, the costs of offloading the mission via ground base stations, and the costs associated with choosing local computing, respectively; O i (h i h -i ) indicates device u i Unloading costs.

[0012] According to a task offloading method for low Earth orbit satellites and ground edge computing networks provided by the present invention, before calculating the cost of each device according to the following formula, the method further includes: determining the cost of task offloading for each device via low Earth orbit satellites based on the total energy consumption of the devices offloading via low Earth orbit satellites; the method for calculating the total energy consumption of the devices offloading via low Earth orbit satellites includes:

[0013]

[0014] Among them, D i It is device u i The amount of task data during the task unloading and transfer process; For device u i The data rate during transmission, where m represents satellite m and n represents the nth channel; For device u i The mission's transmission power is offloaded through the nth channel of satellite m.

[0015] According to a task offloading method for low-Earth orbit satellites and ground edge computing networks provided by the present invention, before calculating the cost of each device according to the following formula, the method further includes: determining the cost of task offloading for each device through the base station based on the total energy consumption of the devices offloaded by the base station; the method for calculating the total energy consumption of the devices offloaded by the base station includes:

[0016]

[0017] Among them, D i It is device u i The amount of task data during the task unloading and transfer process; For device u i The data rate during transmission tasks, where j represents base station j and k represents the kth channel; For device u i The transmit power of the task is offloaded through the k-th channel of base station j.

[0018] According to the present invention, a task offloading method for low Earth orbit satellites and ground edge computing networks, before calculating the cost of each device according to the following formula, further includes: determining the overhead cost of each device in the local computing task based on the device energy consumption generated by local computing;

[0019] The methods for calculating device energy consumption generated by local computing include:

[0020]

[0021] Where ε represents the local computing power factor, which is determined based on the device's own chip architecture; X i This indicates the number of CPU cycles required during the task computation; f i loc Indicates device u i Its own computing power.

[0022] This invention provides a task offloading device for low-Earth orbit satellites and ground edge computing networks, comprising: an initial allocation module, configured to determine the initial decision results of all devices processing tasks locally, offloading them to a base station, or offloading them to low-Earth orbit satellites, as the current state; a cost determination module, configured to calculate the total cost of the system comprised of all devices in the current state; and a first loop module, configured to, for each device, iterate through all decision results of processing tasks locally, offloading them to a base station, or offloading them to low-Earth orbit satellites; determine the decision result that minimizes the total system cost among all decision results for each device, and select a candidate decision result from all decision results of all devices; and make a final decision. The policy update module is used to update the current state based on the candidate decision results if the total system cost corresponding to the candidate decision results is less than the total system cost of the current state. The second loop module is used to repeat the above process of iterating through the decision results for each device until the current state is updated based on the candidate decision results, until the total system cost corresponding to all decision results of each device is not less than the total system cost of the current state. The total system cost is determined based on the cost of all devices. The local computing cost of each device is determined based on the processor resource requirements. The cost of offloading the task to be processed to a base station or a low Earth orbit satellite for processing is determined based on the corresponding transmission energy consumption.

[0023] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the task offloading method for low Earth orbit satellites and ground edge computing networks as described above.

[0024] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the task offloading method for low Earth orbit satellites and ground edge computing networks as described above.

[0025] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the task offloading method for low Earth orbit satellites and ground edge computing networks as described above.

[0026] The task offloading method and apparatus for low-Earth orbit satellites and ground-based edge computing networks provided by this invention can achieve efficient device allocation while reducing the energy consumption generated when ground equipment transmits computing tasks, thus realizing a trade-off between the total energy consumption of devices in the network and device allocation. Furthermore, based on the aforementioned update method, it tends to converge and reach a Nash equilibrium after a certain number of iterations. Therefore, the game-theoretic decision-making method proposed in this invention has superior performance compared to commonly used benchmark methods, and can significantly reduce system energy consumption and improve task offloading efficiency. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0028] Figure 1 This is one of the flowcharts illustrating the task offloading method between low-Earth orbit satellites and ground edge computing networks provided by this invention;

[0029] Figure 2 This is the satellite-ground plan view provided by the present invention;

[0030] Figure 3 This is an application scenario diagram provided by the present invention;

[0031] Figure 4 This is the second flowchart illustrating the task offloading method between low-Earth orbit satellites and ground edge computing networks provided by this invention;

[0032] Figure 5 This is a schematic diagram of the task offloading device for near-Earth orbit satellites and ground edge computing networks provided by the present invention;

[0033] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0035] The following is combined with Figures 1-6 The present invention describes a task offloading method and apparatus for low Earth orbit satellites and ground edge computing networks. Figure 1 This is one of the flowcharts illustrating the task offloading method between low-Earth orbit satellites and ground-based edge computing networks provided by this invention, such as... Figure 1 As shown, this invention provides a task offloading method for low-Earth orbit satellites and ground-based edge computing networks, including:

[0036] 101. Determine the initial decision result for all devices to process pending tasks locally, offload them to base stations, or offload them to low Earth orbit satellites, and use this as the current state.

[0037] Specifically, in 101, the initial state can be either no decision is made for each device or a decision is made randomly.

[0038] 102. Calculate the total cost of the system consisting of all devices in the current state.

[0039] The total system cost is determined based on the cost of all devices; the local computing cost of each device is determined based on the processor resource requirements; and the cost of offloading the tasks to be processed by each device to a base station or a low Earth orbit satellite is determined based on the corresponding transmission energy consumption.

[0040] Processor resource requirements can refer to the number of CPU cycles needed during task computation; however, this is only one implementation method. Offloading the task to be processed to a base station or a low-Earth orbit satellite requires data transmission. The cost for each scenario is determined based on the device's transmission power and transmission duration.

[0041] 103. For each device, iterate through all decision results that allow the task to be processed locally, offloaded to the base station, or offloaded to a low Earth orbit satellite; determine the decision result that minimizes the total system cost among all decision results for each device, and select a candidate decision result from all decision results for all devices.

[0042] 104. If the total system cost corresponding to the candidate decision result is less than the total system cost of the current state, then the current state is updated according to the candidate decision result.

[0043] 105. Repeat the above process for each device, iterating through the decision results until the current state is updated based on the candidate decision results, until the total system cost corresponding to all decision results for each device is not less than the total system cost of the current state.

[0044] Steps 101 to 105 above are related. To make the steps of this invention clearer, the steps will now be described as a whole below. Before introducing the low Earth orbit satellite-ground edge computing network, a visual plan view of the relationship between the low Earth orbit satellite and ground equipment (as shown in the attached diagram) is needed. Figure 2 (As shown) to understand the communication and motion trends of low Earth orbit satellites. The elevation angle between the mobile device and the low Earth orbit satellite can be equated to angle β, which can also be expressed as...

[0045]

[0046] Depend on Furthermore, we can obtain:

[0047]

[0048] The Earth's center passes through the line y1 connecting the point mass and the near-Earth orbit satellite, and the Earth's center passes through the line y2 connecting the mobile device and the farthest moving position of the near-Earth orbit satellite. The angle between y1 and y2 can be expressed as:

[0049]

[0050] The arc length G of the trajectory of a near-Earth orbit satellite to its farthest position can be expressed as:

[0051] G=2α(Re+He)#(4)

[0052] The velocity of the near-Earth orbit satellite is v leo From the above, we can see the maximum communication time that mobile devices can use for offloading via low Earth orbit satellites. It can be represented as:

[0053]

[0054] Mobile devices u i and low Earth orbit satellites m The distance between them can be represented as d i,m :

[0055]

[0056] This invention combines low-Earth orbit (LEO) satellite networks and terrestrial communication systems, and introduces mobile edge computing to decentralize abundant computing resources to the edge of LEO satellite networks and terrestrial base stations. Furthermore, it designs an energy-efficient device allocation and task offloading method, reducing data transmission energy consumption for device requesting offloading services under the constraints of satellite communication latency and device elevation angle.

[0057] The near-Earth orbit satellite-ground edge computing network proposed in this invention considers the device task offloading problem under the cross-coverage of multiple base stations and multiple devices. The application scenario model is as follows: Figure 3 As shown, low Earth orbit satellites and ground base stations are deployed to provide mission offloading services, using a low Earth orbit satellite set L = {l1, l2, ..., l...} v} and the base station set S = {s1, s2, ..., s} w} represents the computing resources required for the device to request the offloading service, and the set of ground devices is represented as U = {u1, u2, ..., u}. x In this network, any device u i It may be within the coverage area of ​​multiple base stations, and these base stations may have overlapping network coverage areas. In other words, any device can have multiple base stations that satisfy the proximity constraint (15) to choose from for requesting services. Any ground device u i The computational task when requesting data services is represented by K. i ={D i ,X i}, where D i X is the size of the task data during the task unloading and transfer process. i This indicates the number of CPU cycles required during the task computation. Additionally, f... i loc Indicates device u i Each device has its own computing power, which may vary.

[0058] In a low Earth orbit satellite-ground edge computing network, there are three offloading computation methods for device request tasks: low Earth orbit satellite offloading, ground base station offloading, and local computation. The offloading decision variable for each device is h. i =(a i b i e i ), where a i The decision is based on the offloading method (low Earth orbit satellite, ground base station, or local on the device), b i Indicates server selection decision (l) m or s j (where m represents satellite m, j represents base station j), e i Represents the channel selection decision (c m，n or cj，k (where n represents the nth channel of the satellite and k represents the kth channel of the base station), the decision set of all devices is H = {h1, h2, ..., hk}. x In addition, device u i When no decision is made, it can be represented as

[0059] In this invention, each device in the system aims to minimize its own cost, which is also the optimization objective of this invention: minimizing the total cost of all devices. Of course, some corresponding constraints are also included.

[0060] The optimization problem of minimizing the total cost of all devices is a stochastic optimization problem, and obtaining an NP-hard solution is extremely difficult. This invention solves this optimization problem using potential game theory. Potential game theory is a game theory tool for solving competition between devices under multiple constraints, suitable for solving competition problems and minimizing the overall cost of the model. Therefore, the corresponding model can be expressed as a non-cooperative competitive game model G between devices for solution.

[0061] In model G, due to the principle of individual rationality among devices, given limited system resources, each device tends to reduce its own cost. After multiple iterations, the system will reach a Nash equilibrium. In this state, no device can unilaterally change its own strategy (while the strategies of other devices remain unchanged) without reducing its own cost. That is, the total system cost corresponding to all decision outcomes of each device in model G is no less than the total system cost of the current state.

[0062] To address the issue of device task offloading (choosing which offloading method to use), this invention proposes the aforementioned task offloading algorithm based on game theory. This algorithm iteratively updates the device's decision by considering the different costs associated with choosing different offloading methods, thereby minimizing the total system cost.

[0063] In the algorithm of this invention, all devices undergo a finite number of iterations. Each device updates its strategy based on its own minimum cost until no device is willing to update its strategy to change its cost, thus reaching a Nash equilibrium state. The specific process is as follows: Figure 4 As shown in Figure 103, a candidate decision result is selected from the decision results of all devices. This selection can be random or as shown in Figure 103. Figure 4 In competitive selection, or rather Figure 4 The competitive selection process can be achieved through random selection.

[0064] The task offloading method for low-Earth orbit satellites and ground-based edge computing networks of this invention can achieve efficient device allocation while reducing the energy consumption generated when ground equipment transmits computing tasks, thus realizing a trade-off between total energy consumption and device allocation in the network. Furthermore, based on the aforementioned update method, it tends to converge and reach a Nash equilibrium after a certain number of iterations. Therefore, the game-theoretic decision-making method proposed in this invention has superior performance compared to commonly used benchmark methods, and can significantly reduce system energy consumption and improve task offloading efficiency.

[0065] In one embodiment, the intra-cell interference caused by other devices on the same channel to any device selecting a low-Earth orbit satellite for task offloading is less than a preset threshold; the total delay of any device selecting a low-Earth orbit satellite for task offloading is less than a preset threshold; the intra-cell power interference of any device to all devices on the same base station and its channel is less than a preset threshold; and the total computing power provided by each base station to devices within its coverage area is less than the maximum computing power of each base station. See the constraints in (8), (10), (14), (15), and (17) below for details.

[0066] Equipment selection: Low Earth orbit satellite m The task is offloaded from n channels, i.e., its decision variable is h. i =(a i b i e i ) = (1, m, n).

[0067] Data rate when the device UI transmits tasks Represented as:

[0068]

[0069] Equipment u i Subjected to the same channel c m，n Interference caused by other equipment within the cell must meet the following requirements:

[0070]

[0071] in, Indicates device u t The channel gain of the offloading task is achieved through the nth channel of server m. Indicates device u t The transmit power Q when selecting the nth channel of server m i This indicates a preset threshold.

[0072] Therefore, the total delay for the equipment to select a low-Earth orbit satellite for mission unloading can be expressed as:

[0073]

[0074] It consists of three parts: the propagation delay caused by the vacuum medium. Transmission delay and computational delay c is the speed of light, f i，m The computing power allocated to the device for the low Earth orbit satellite server. Device u i via low Earth orbit satellite m The total energy consumption of the device during unloading is:

[0075]

[0076] The unloading cost in this state can be expressed as:

[0077]

[0078] When device u i The uninstallation decision is h i =(a i b i e i ) = (2, j, k), meaning the equipment selects ground base station s. j k th Each channel offloads its own computational tasks, and the data rate during offloading transmission tasks is:

[0079]

[0080] At this time, device u i Will be subject to selection of the same base station s j and its channel c j，k The intra-cell power interference of all devices, and this interference must meet the following requirements:

[0081]

[0082] θ j Indicates base station s j The radius of the coverage area, q i，j Indicates ground equipment u i and base station s j The distance between them. Therefore, at base station s j The set of devices within the coverage area can be represented as:

[0083] S i ={s j ∈S|d i，j ≤θ j}#(15)

[0084] Equipment u i via base station s j The total latency during uninstallation is:

[0085]

[0086] It consists of two parts: the transmission delay for selective base station offloading. And calculation delay during unloading The following constraints must also be met:

[0087]

[0088] f i，j F represents the computing power allocated to the equipment by the base station server. j For base station s j Maximum computing power of device u. i via base station s j The total energy consumption during unloading is:

[0089]

[0090] The unloading cost in this state is:

[0091]

[0092] Based on the above, each device in the system aims to minimize its own cost, which is also the optimization goal of this invention: to minimize the total cost of all devices.

[0093] In one embodiment, the cost of each device is also calculated according to the following formula:

[0094]

[0095]

[0096] Among them, h i For device u i The resulting decision, h -i Indicates excluding u i Decisions for all other equipment; C i (h i h -i ) indicates device u i Uninstallation decision h i The resulting uninstallation costs, These represent the costs of offloading the mission via low Earth orbit satellites, the costs of offloading the mission via ground base stations, and the costs associated with choosing local computing, respectively; O i (h i h -i ) indicates device u i Unloading costs.

[0097] Equipment u i The decision to perform the computation task locally, i.e., its offloading decision is h. i =(a i b i e i The computational latency of a task on a local computing device can be expressed as: (0, 0, 0).

[0098]

[0099] Meanwhile, the energy consumption of the device generated by local computing is:

[0100]

[0101] Where ε represents the local computing energy consumption factor, determined based on the device's own chip architecture. The corresponding offloading cost for the device is then expressed as:

[0102]

[0103] except u i The decision of all other devices is represented by h. -i Then it is determined by device u i Uninstallation decision h i The resulting unloading cost is expressed as follows:

[0104]

[0105] When device u i When h abandons its uninstallation decision -t，i Let u represent the set of decisions made by all other devices. Consider the impact on device u when the decisions of other devices remain constant and when they change. i The impact of unloading costs will affect the device u i Uninstallation decision h i The resulting unloading cost is expressed as:

[0106]

[0107] Based on the above, each device in this system aims to minimize its own cost, which is also the optimization objective of this invention: minimizing the total cost of all devices.

[0108]

[0109] st(8),(10),(14),(15),(17)#(25)

[0110] In one embodiment, before calculating the cost of each device according to the following formula, the method further includes: determining the cost of unloading the mission via a low-Earth orbit satellite based on the total energy consumption of the equipment unloaded from the low-Earth orbit satellite; the method for calculating the total energy consumption of the equipment unloaded from the low-Earth orbit satellite includes:

[0111]

[0112] Among them, D i It is device u i The amount of task data during the task unloading and transfer process; For device u i The data rate during transmission, where m represents satellite m and n represents the nth channel; For device u i The transmit power of the task is offloaded through the nth channel of base station m.

[0113] Then, based on the above formula (12), the cost of unloading the mission for each device via a low Earth orbit satellite can be obtained, as described above.

[0114] In one embodiment, before calculating the cost of each device according to the following formula, the method further includes: determining the cost of offloading tasks through the base station for each device based on the total energy consumption of the devices offloaded from the base station; the method for calculating the total energy consumption of the devices offloaded from the base station includes:

[0115]

[0116] Among them, D i It is device u i The amount of task data during the task unloading and transfer process; For device u i The data rate during transmission tasks, where j represents base station j and k represents the kth channel; For device u i The transmit power of the task is offloaded through the k-th channel of base station j.

[0117] Then, based on the above formula (19), the cost of task offloading for each device through the base station can be obtained, as can be found in the above content.

[0118] In one embodiment, before calculating the cost of each device according to the following formula, the method further includes: determining the overhead cost of each device in the local computing task based on the device energy consumption generated by the local computing; the method for calculating the device energy consumption generated by the local computing includes:

[0119]

[0120] Where ε represents the local computing power consumption factor, which is determined based on the device's own chip architecture; fi loc ;X i This indicates the number of CPU cycles required during the task computation; f i loc Indicates device u i Its own computing power.

[0121] Then, based on the above formula (22), the overhead cost of each device's local computing task can be obtained, as can be found in the above content.

[0122] The advantages and feasibility of the method of the present invention will now be further explained. In the above embodiments, model G can be defined as:

[0123]

[0124] In the model In this system, due to the principle of individual rationality among devices, given limited system resources, each device tends to reduce its own cost. After multiple iterations, the system will reach a Nash equilibrium. In this state, no device can reduce its own cost by unilaterally changing its strategy (while the strategies of other devices remain unchanged).

[0125] Nash equilibrium is defined as follows: In this non-cooperative competitive game model... In the middle, if there exists a decision set for all devices... In this state, no single device will unilaterally change its decision to reduce costs.

[0126]

[0127] Definition of potential game: In this model In the context, for any device u i ∈U, given the device u i The decision set h of all other devices -i When there exists a potential function satisfy:

[0128]

[0129] Model It is a perfect power game.

[0130] It is a complete potential game, and the potential function can be expressed as:

[0131]

[0132] Regarding the performance of this model, this invention theoretically proposes a PoA (Price of Anarchy) for allocating the number of users and the total system cost. The advantages of this model and algorithm are shown through comparison with Nash equilibrium, and the limits of PoA are also proved.

[0133] (1) PoA of the number of users allocated to low Earth orbit satellites

[0134] Formula (7) can be used to express the product of transmit power and channel gain when the equipment selects satellite unloading as follows: We can obtain:

[0135]

[0136] It is the collection of all equipment allocated to low Earth orbit satellites. It is the number of all elements in the set, that is, the number of elements allocated to satellite l. m The number of devices on it. It is the set of all equipment distributed on low Earth orbit satellites under Nash equilibrium conditions. This corresponds to the number of devices. Meanwhile, the PoA for allocating users to low Earth orbit satellites in this model can be defined as:

[0137]

[0138] According to formula (8), we can know that:

[0139] Q min =min{Q i}, Q max =max{Q i}#(32)

[0140] The range of values ​​for the PoA (Pocket Aptitude) for the number of users allocated to low Earth orbit satellites satisfies:

[0141]

[0142] (2) PoA of the number of users allocated by the ground base station

[0143] The product of transmit power and channel gain when the device selects base station offloading is expressed as: The relevant definition is expressed as follows:

[0144]

[0145] It is the collection of all devices allocated to ground base stations. It is the number of all elements in the set, that is, the number of elements allocated to base station l. m The number of devices on it. It is the set of all devices distributed on ground base stations under Nash equilibrium. This refers to the corresponding number of devices. Meanwhile, the PoA for user allocation at ground base stations in this model can be defined as:

[0146]

[0147] According to formula (14), we can obtain:

[0148] R min =min{R i}, R max =max{R i}#(36)

[0149] The PoA limit for the number of users allocated by the terrestrial base station satisfies:

[0150]

[0151] According to formula (27) and the general definition of PoA, the PoA of the total cost of the system is expressed as:

[0152]

[0153] For making the uninstallation decision h i device u i Its maximum cost can be expressed as Ma(h) i )=max{O i (h i h -i The minimum cost is expressed as Ms(h) i )=min{O i (h i h -i Therefore, by using formulas (23), (24) and (38), we can obtain the range of values ​​for the total cost PoA.

[0154] The range of values ​​for the total cost PoA satisfies:

[0155]

[0156] Therefore, this invention tends to converge and reach Nash equilibrium after a certain number of iterations. Thus, the game-theoretic decision-making method proposed in this invention can significantly reduce system energy consumption and improve task offloading efficiency.

[0157] The task offloading device for low Earth orbit satellites and ground edge computing networks provided by the present invention is described below. The task offloading device for low Earth orbit satellites and ground edge computing networks described below can be referred to in correspondence with the task offloading method for low Earth orbit satellites and ground edge computing networks described above.

[0158] Figure 5 This is a schematic diagram of the task offloading device for low-Earth orbit satellites and ground edge computing networks provided by the present invention, as shown below. Figure 5 As shown, the task offloading device for the low Earth orbit satellite and the ground edge computing network includes: an initial allocation module 501, a cost determination module 502, a first loop module 503, a decision update module 504, and a second loop module 505. The initial allocation module 501 determines the initial decision result for all devices to process the task locally, offload it to the base station, or offload it to the low Earth orbit satellite, as the current state. The cost determination module 502 calculates the total cost of the system comprised of all devices in the current state. The first loop module 503 iterates through all decision results for each device, determining the decision result that minimizes the total system cost among all decision results for each device, and selects a candidate decision result from all device decision results. The decision update module 504 updates the decision result if the candidate decision result is selected. If the total system cost corresponding to the candidate decision result is less than the total system cost of the current state, then the current state is updated according to the candidate decision result. The second loop module 505 is used to repeat the above process of iterating through the decision results for each device until the current state is updated according to the candidate decision results, until the total system cost corresponding to all decision results of each device is not less than the total system cost of the current state. The total system cost is determined based on the cost of all devices. The local computing cost of each device is determined based on the processor resource requirements. The cost of offloading the task to be processed to the base station or low Earth orbit satellite for processing is determined based on the corresponding transmission energy consumption.

[0159] The apparatus embodiments provided in this invention are for implementing the above-described method embodiments. For specific processes and details, please refer to the above-described method embodiments, which will not be repeated here.

[0160] The task offloading device for low-Earth orbit satellites and ground edge computing networks provided in this embodiment of the invention has the same implementation principle and technical effects as the aforementioned task offloading method embodiment for low-Earth orbit satellites and ground edge computing networks. For the sake of brevity, any parts not mentioned in the embodiment of the task offloading device for low-Earth orbit satellites and ground edge computing networks can be referred to the corresponding content in the aforementioned task offloading method embodiment for low-Earth orbit satellites and ground edge computing networks.

[0161] Figure 6 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 6As shown, the electronic device may include: a processor 601, a communication interface 602, a memory 603, and a communication bus 604. The processor 601, communication interface 602, and memory 603 communicate with each other via the communication bus 604. The processor 601 can call logical instructions in the memory 603 to execute a task offloading method between the low Earth orbit satellite and the ground edge computing network. This method includes: determining the initial decision results of all devices processing the task locally, offloading it to the base station, or offloading it to the low Earth orbit satellite, as the current state; calculating the total cost of the system comprised of all devices in the current state; for each device, iterating through all decision results of processing the task locally, offloading it to the base station, or offloading it to the low Earth orbit satellite; determining the decision result that minimizes the total system cost among all decision results for each device, and selecting a candidate decision result from all device decision results. Select a decision result; if the total system cost corresponding to the candidate decision result is less than the total system cost of the current state, then update the current state according to the candidate decision result; repeat the above process of iterating through the decision results for each device until the current state is updated according to the candidate decision result, until the total system cost corresponding to all decision results of each device is not less than the total system cost of the current state; wherein, the total system cost is determined based on the cost of all devices; the local computing cost of each device is determined based on the processor resource requirements; the cost for each device to offload the task to be processed to the base station for processing or the low Earth orbit satellite for processing is determined based on the corresponding transmission energy consumption.

[0162] Furthermore, the logical instructions in the aforementioned memory 603 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0163] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the task offloading method for low Earth orbit satellites and ground edge computing networks provided by the above methods. The method includes: determining the initial decision results of all devices processing the task to be processed locally, offloading it to a base station for processing, or offloading it to a low Earth orbit satellite for processing, as the current state; calculating the total cost of all devices constituting the system in the current state; for each device, iterating through all decision results of processing the task to be processed locally, offloading it to a base station for processing, or offloading it to a low Earth orbit satellite for processing; and determining all decision results of each device. The system selects the decision that minimizes the total system cost from all device decisions, and then selects a candidate decision from all device decisions. If the total system cost corresponding to the candidate decision is less than the total system cost of the current state, the current state is updated based on the candidate decision. This process is repeated for each device, iterating through the decision results until the current state is updated based on the candidate decision, until the total system cost corresponding to all decisions for each device is not less than the total system cost of the current state. The total system cost is determined based on the cost of all devices. The local computing cost of each device is determined based on the processor resource requirements. The cost for each device to offload the task to be processed to a base station or a low Earth orbit satellite is determined based on the corresponding transmission energy consumption.

[0164] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements a task offloading method for low Earth orbit satellites and ground edge computing networks provided by the methods described above. This method includes: determining an initial decision result for all devices to process the task to be processed locally, offload it to a base station, or offload it to a low Earth orbit satellite, as the current state; calculating the total cost of the system comprised of all devices in the current state; for each device, iterating through all decision results for processing the task to be processed locally, offloading it to a base station, or offloading it to a low Earth orbit satellite; and determining the decision result that minimizes the total system cost among all decision results for each device. The system selects a candidate decision from all the decision results of all devices. If the total system cost corresponding to the candidate decision is less than the total system cost of the current state, the current state is updated according to the candidate decision. This process of iterating through the decision results for each device and updating the current state according to the candidate decision is repeated until the total system cost corresponding to all decision results of each device is not less than the total system cost of the current state. The total system cost is determined based on the cost of all devices. The local computing cost of each device is determined based on the processor resource requirements. The cost for each device to offload the task to be processed to a base station or a low Earth orbit satellite is determined based on the corresponding transmission energy consumption.

[0165] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0166] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for offloading tasks between low-Earth orbit satellites and ground-based edge computing networks, characterized in that, include: The initial decision result for all devices to process pending tasks locally, offload them to base stations, or offload them to low Earth orbit satellites is determined as the current state; Calculate the total cost of the system comprised of all devices in the current state; For each device, iterate through all decision results that allow the task to be processed locally, offloaded to the base station for processing, or offloaded to a low Earth orbit satellite for processing. Determine the decision outcome that minimizes the total system cost among all decision outcomes for each device, and select a candidate decision outcome from all decision outcomes for all devices; If the total system cost corresponding to the candidate decision result is less than the total system cost of the current state, then the current state is updated according to the candidate decision result; Repeat the above process for each device, iterating through the decision results until the current state is updated based on the candidate decision results, until the total system cost corresponding to all decision results for each device is not less than the total system cost of the current state; The total system cost is determined based on the cost of all devices; the local computing cost of each device is determined based on the processor resource requirements; and the cost of offloading the tasks to be processed by each device to a base station or a low Earth orbit satellite is determined based on the corresponding transmission energy consumption. The decision result of each device corresponds to its unloading decision variable, which includes unloading method decision, server selection decision, and channel selection decision. If any device selects a low Earth orbit satellite for mission offloading, the intra-cell interference caused by other devices on the same channel is less than a preset threshold. The total delay for any device to select a low Earth orbit satellite for mission unloading is less than a preset threshold. The intra-cell power interference of any device from all devices on the same base station and its channel is less than a preset threshold. The total computing power provided by each base station to devices within its coverage area is less than the maximum computing power of each base station. It also includes calculating the cost of each device using the following formula: ； ； in, For equipment The resulting decisions ,in, The decision is based on the uninstallation method. This indicates the server selection decision. This represents the channel selection decision; m represents satellite m; n represents the nth channel of the satellite; j represents base station j; k represents the kth channel of the base station. Indicates except Decisions regarding all other equipment; Indicates equipment Uninstallation decision The resulting uninstallation costs, , , These represent the costs of offloading the mission via low Earth orbit satellites, the costs of offloading the mission via ground base stations, and the costs incurred by choosing local computing, respectively. Indicates equipment Unloading costs; Indicates when the device The set of decisions for all other devices when abandoning its offloading decision; The total cost of the system is expressed as: ； The system is optimized with the goal of minimizing the total system cost; wherein, This refers to a collection of ground equipment.

2. The task offloading method for low-Earth orbit satellites and ground edge computing networks according to claim 1, characterized in that, Before calculating the cost of each device according to the following formula, the process also includes: Determine the cost of unloading each piece of equipment via a low Earth orbit satellite based on the total energy consumption of the equipment being unloaded from the low Earth orbit satellite. The method for calculating the total energy consumption of equipment unloading from the near-Earth orbit satellite includes: ； in, It is equipment The amount of task data during the task unloading and transfer process; For equipment Data rate during task transmission Indicates satellite , Indicates the first Channel; For equipment The mission's transmission power is offloaded through the nth channel of satellite m.

3. The task offloading method for low-Earth orbit satellites and ground edge computing networks according to claim 1, characterized in that, Before calculating the cost of each device according to the following formula, the process also includes: The cost of offloading tasks for each device through the base station is determined based on the total energy consumption of the devices offloaded from the base station. The method for calculating the total energy consumption of the equipment offloaded from the base station includes: ； in, It is equipment The amount of task data during the task unloading and transfer process; For equipment Data rate during task transmission Indicates base station , Indicates the first Channel; For equipment via base station The The transmit power of the offloading task on each channel.

4. The task offloading method for low-Earth orbit satellites and ground edge computing networks according to claim 1, characterized in that, Before calculating the cost of each device according to the following formula, the process also includes: Determine the overhead cost of each device for local computing tasks based on the device energy consumption generated by local computing. The device energy consumption calculation method generated by the local calculation includes: ； in, This represents the local computing power consumption factor, which is determined based on the device's own chip architecture. This indicates the number of CPU cycles required during the task calculation process; Indicates equipment Its own computing power.

5. A task offloading device for low-Earth orbit satellites and ground-based edge computing networks, characterized in that, include: The initial allocation module is used to determine the initial decision result of all devices whether to process the pending tasks locally, offload them to the base station for processing, or offload them to the low Earth orbit satellite for processing, and this is used as the current state. The cost determination module is used to calculate the total cost of the system comprised of all devices in the current state. The first loop module is used to iterate through all decision results for each device, whether to process the task locally, offload it to the base station, or offload it to a low Earth orbit satellite; determine the decision result that minimizes the total system cost among all decision results for each device, and select a candidate decision result from all decision results of all devices; The decision update module is used to update the current state based on the candidate decision result if the total system cost corresponding to the candidate decision result is less than the total system cost of the current state. The second loop module is used to repeat the above process of iterating through the decision results for each device until the current state is updated based on the candidate decision results, until the total system cost corresponding to all decision results of each device is not less than the total system cost of the current state. The total system cost is determined based on the cost of all devices; the local computing cost of each device is determined based on the processor resource requirements; and the cost of offloading the tasks to be processed by each device to the base station or low Earth orbit satellite is determined based on the corresponding transmission energy consumption. The decision result of each device corresponds to its unloading decision variable, which includes unloading method decision, server selection decision, and channel selection decision. If any device selects a low Earth orbit satellite for mission offloading, the intra-cell interference caused by other devices on the same channel is less than a preset threshold. The total delay for any device to select a low Earth orbit satellite for mission unloading is less than a preset threshold. The intra-cell power interference of any device from all devices on the same base station and its channel is less than a preset threshold. The total computing power provided by each base station to devices within its coverage area is less than the maximum computing power of each base station. It also includes calculating the cost of each device using the following formula: ； ； in, For equipment The resulting decisions ,in, The decision is based on the uninstallation method. This indicates the server selection decision. This represents the channel selection decision; m represents satellite m; n represents the nth channel of the satellite; j represents base station j; k represents the kth channel of the base station. Indicates except Decisions regarding all other equipment; Indicates equipment Uninstallation decision The resulting uninstallation costs, , , These represent the costs of offloading the mission via low Earth orbit satellites, the costs of offloading the mission via ground base stations, and the costs incurred by choosing local computing, respectively. Indicates equipment Unloading costs; Indicates when the device The set of decisions for all other devices when abandoning its offloading decision; The total cost of the system is expressed as: ； The system is optimized with the goal of minimizing the total system cost; wherein, This refers to a collection of ground equipment.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the task offloading method for low Earth orbit satellites and ground edge computing networks as described in any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the task offloading method for low Earth orbit satellites and ground edge computing networks as described in any one of claims 1 to 4.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the task offloading method for low Earth orbit satellites and ground edge computing networks as described in any one of claims 1 to 4.

Images