Task Offloading and Resource Optimization Method and System for 5G and Satellite Converged Network

Abstract

The present invention discloses a task offloading and resource optimization method and system for a 5G and satellite integrated network, which concretizes a pre-constructed system model into a task segmentation model, a data transmission model, a computing model, a system delay constraint model, and an energy consumption model; uses the task segmentation model, the data transmission model, the computing model, the system delay constraint model, and the energy consumption model to establish a system model optimization problem, transforms the long-term average energy consumption optimization problem of power transmission and distribution equipment into an energy consumption optimization problem within a single time slot, decouples the energy consumption optimization problem within a single time slot into three separate sub-problems, and sequentially solves the three separate sub-problems to obtain an optimal solution that can minimize the long-term average energy consumption of power transmission and distribution equipment. The present invention can provide on-demand computing capabilities for power transmission and distribution equipment, minimize the long-term average energy consumption of power transmission and distribution equipment, and improve network performance.

Description

Technical Field

[0001] The present invention belongs to the technical field of 5G and satellite integrated networks, and particularly relates to a task offloading and resource optimization method and system for a 5G and satellite integrated network for the power transmission and distribution scenario. Background Art

[0002] Power transmission and distribution equipment equipped with advanced embedded monitoring and data acquisition technologies plays an important role in numerous power applications and services. With the rapid development of power transmission and distribution services such as drone inspection, precise load control, and power metering, each piece of power transmission and distribution equipment will generate a large number of computationally intensive tasks, which poses a huge challenge to the computing capabilities of resource-constrained power Internet of Things devices.

[0003] The 5G and satellite integrated network combining 5G communication technology and satellite technology is an important way to solve the above problems. The 5G network, based on technologies such as software-defined and network function virtualization, can support on-demand customization, high-dynamic expansion, and automated deployment of network resources. Applying 5G communication technology to the power transmission and distribution scenario can further improve the system performance of the wireless private network and enhance the multi-service differentiated security bearing capacity. At the same time, in recent years, the development of Internet satellite constellations has been rapid. It can provide services such as navigation and positioning, precise timekeeping, and short message communication for the construction of the power transmission and distribution scenario, and mostly uses Ka or Ku bands, which greatly improves the system capacity and can provide high-speed broadband Internet access services for areas where the cost of traditional Internet infrastructure is too high, making up for the deficiencies of ground base stations or insufficient resources. It is an effective supporting means for the strategic development of China's energy. Therefore, integrating 5G networks and satellite networks into the power transmission and distribution scenario can provide on-demand computing capabilities for power transmission and distribution equipment through ground 5G base stations and geostationary earth orbit (GEO) satellites, thereby improving the utility of the system and reducing the energy consumption of power transmission and distribution equipment.

[0004] However, the actual 5G and satellite integrated network for the power transmission and distribution scenario operates in a random dynamic environment. Therefore, the task offloading and resource optimization of the 5G and satellite integrated network for the power transmission and distribution scenario still face several challenging problems. First, the dynamic behaviors such as the arrival, transmission, and processing of tasks in the 5G and satellite integrated network are random processes over a period of time, and the long-term performance of the system needs to be considered. Second, the transmission energy consumption of offloading tasks from power transmission and distribution equipment to ground 5G base stations and GEO satellites is different. Therefore, tasks need to be reasonably allocated to extend the service life of the equipment. Finally, ground 5G base stations and GEO satellites have different computing resources and are subject to different constraints, and the computing resources under dynamic network conditions need to be efficiently allocated. Summary of the Invention

[0005] The object of the present invention is to provide a task offloading and resource optimization method and system for a 5G and satellite integrated network for the power transmission and distribution scenario, so as to overcome the defects existing in the prior art. The present invention can provide on-demand computing capabilities for power transmission and distribution equipment, minimize the long-term average energy consumption of power transmission and distribution equipment, and improve network performance.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A task offloading and resource optimization method for a 5G and satellite integrated network, including:

[0008] Specify the pre-constructed system model as a task segmentation model, a data transmission model, a computing model, a system delay constraint model, and an energy consumption model;

[0009] Establish a system model optimization problem by using the task segmentation model, the data transmission model, the computing model, the system delay constraint model, and the energy consumption model; wherein, the total optimization time of the system model is divided into T equal time slots;

[0010] Convert the long-term average energy consumption optimization problem of the power transmission and distribution equipment into an energy consumption optimization problem within a single time slot;

[0011] Decouple the energy consumption optimization problem within the single time slot into at least one separate sub-problem, and sequentially solve the sub-problems to obtain an optimal solution that can minimize the long-term average energy consumption of the power transmission and distribution equipment.

[0012] Furthermore, the pre-constructed system model includes a geostationary satellite GEO, N 5G base stations, and M power transmission and distribution equipment. The task segmentation model is used to divide the A m (t) units of computing tasks to be processed generated by the power transmission and distribution equipment m in each time slot into Tasks for local computing, Tasks offloaded to the 5G base station, Tasks offloaded to the geostationary satellite GEO, and the tasks for local computing, the tasks offloaded to the 5G base station, and the tasks offloaded to the geostationary satellite GEO are executed in parallel;

[0013] The power transmission and distribution equipment m is provided with buffer queues for storing the tasks for local computing, the tasks offloaded to the 5G base station, and the tasks offloaded to the geostationary satellite GEO respectively and

[0014] Furthermore, the data transmission model is used to calculate the data transmission rate, specifically: each power transmission and distribution equipment is equipped with two communication interfaces for communicating with the 5G base station and the geostationary satellite GEO respectively;

[0015] When the power transmission and distribution equipment m has a task to be transmitted to the nth 5G base station for calculation at time slot t, the data transmission rate R is obtained according to Shannon's formula. m When the power transmission and distribution equipment m has a task to be transmitted to the geostationary satellite GEO for calculation at time slot t, a fixed data transmission rate is adopted.

[0016] Furthermore, the calculation model is used to calculate the amount of task data completed in the current time slot. Specifically:

[0017] At time slot t, when a task is selected to be calculated locally, the local computing resources are used to calculate the amount of task data that can be completed in the current time slot.

[0018] At time slot t, when a task is selected to be offloaded to the nth 5G base station for calculation, the queue backlog W m,n (t) after the task is transmitted to the nth 5G base station is considered, and then the amount of task data completed is obtained by optimizing the computing resources allocated by the 5G base station for the power transmission and distribution equipment.

[0019] At time slot t, when a task is selected to be offloaded to the geostationary satellite GEO for calculation, the amount of task data completed is obtained by multiplying the time slot length by the data transmission rate from the ground to the satellite.

[0020] Furthermore, the system delay constraint model ensures the average delay of the system by restricting the average time length of the task queue, and the task queue includes the buffer queue of the task and the backlog queue of the task transmitted to the 5G base station.

[0021] Furthermore, the energy consumption model is used to calculate the sum of the local computing energy consumption of each power transmission and distribution equipment, the energy consumption of task offloading from the power transmission and distribution equipment to the 5G base station, and the energy consumption of task offloading from the power transmission and distribution equipment to the geostationary satellite GEO.

[0022] Furthermore, the system model optimization problem is specifically the long-term average energy consumption optimization problem of the power transmission and distribution equipment, which is expressed as:

[0023] P1:

[0024] s.t.C 1 :

[0025] C 2 :

[0026] C 3 :

[0027] C 4 :

[0028] C5 :

[0029] C 6 :

[0030] Among them, the optimized variables are the task segmentation A(t) of the power transmission and distribution equipment, A(t) = {A m (t), m ∈ M}, the task offloading decision a(t) and the computing resource allocation f n (t) of the 5G base station. C 1 represents the task allocation limit of each power transmission and distribution equipment. C 2 represents that the computing power allocated to the power transmission and distribution equipment is limited by the available computing resources of the 5G base station. C 3 and C 4 represent that each power transmission and distribution equipment can select at most one base station for connection in each time slot. C 5 ensures the delay constraint of the system. C 6 represents the non-negativity of the optimized variables. M represents the set of power transmission and distribution equipment, N represents the set of 5G base stations, and when the total optimization time T = {1,..., t,... T} is divided into T time slots, an equal time slot model is adopted.

[0031] Furthermore, the energy consumption optimization problem within the single time slot is decoupled into at least one separate sub-problem, specifically: the energy consumption optimization problem within the single time slot is decoupled into three independent sub-problems, namely the power transmission and distribution equipment task segmentation problem SP1, the power transmission and distribution equipment and 5G base station association control optimization problem SP2, and the 5G base station computing resource allocation problem SP3.

[0032] Furthermore, the power transmission and distribution equipment and 5G base station association control optimization problem SP2 is realized through a many-to-one matching game, specifically including:

[0033] Initialize the set of power transmission and distribution equipment, the set of 5G base stations, and the preference lists of each power transmission and distribution equipment and 5G base station;

[0034] The matching process starts from the power transmission and distribution equipment, and selects the most preferred 5G base station n in the priority list P m (t) of the power transmission and distribution equipment m;

[0035] If the selected 5G base station n has available slots, add the matching pair (m, n) to the matching set. If the 5G base station n has no available slots, compare the power transmission and distribution equipment m with all the power transmission and distribution equipment currently matched with the 5G base station n. If the power transmission and distribution equipment m is better than the worst-matched power transmission and distribution equipment m', then exchange the power transmission and distribution equipment m' and the power transmission and distribution equipment m;

[0036] When there is no blocking pair or the preference lists of all power transmission and distribution equipment and 5G base stations are empty, the matching terminates, and a stable matching between the power transmission and distribution equipment and the 5G base stations at time slot t is obtained, thus obtaining the optimal solution of the optimal control problem SP2 for the association between the power transmission and distribution equipment and the 5G base stations.

[0037] Furthermore, the solution of the 5G base station computing resource allocation problem SP3 adopts a 5G base station computing resource allocation strategy based on minimizing the objective function with queue awareness. The specific process is as follows:

[0038] In each 5G base station, the computing task queue with the longest queue is given priority, and the computing resource requirements of the computing task queue with the longest queue in each 5G base station are satisfied to obtain a computing resource allocation scheme based on queue awareness;

[0039] Initialize the computing resources of each 5G base station. In each 5G base station, the computing task queue that can achieve the minimum SP3 objective function value is given priority, and its computing resource requirements are satisfied to obtain a computing resource allocation scheme based on minimizing the objective function;

[0040] According to the weight factors of the computing resource allocation of the above two computing resource allocation schemes, the optimal solution of the 5G base station computing resource allocation problem SP3 is obtained, thereby minimizing the long-term average energy consumption of the power transmission and distribution equipment.

[0041] The task offloading and resource optimization system for the 5G and satellite integrated network includes:

[0042] System model concretization module: used to concretize the pre-constructed system model into a task segmentation model, a data transmission model, a computing model, a system delay constraint model, and an energy consumption model;

[0043] System model optimization problem establishment module: used to establish a system model optimization problem by using the task segmentation model, the data transmission model, the computing model, the system delay constraint model, and the energy consumption model; among them, the total optimization time of the system model is divided into T equal time slots;

[0044] Optimization problem transformation module: used to transform the long-term average energy consumption optimization problem of the power transmission and distribution equipment into an energy consumption optimization problem within a single time slot;

[0045] Optimal solution solving module: used to decouple the energy consumption optimization problem within a single time slot into at least one separate sub-problem, and sequentially solve the sub-problems to obtain an optimal solution that can minimize the long-term average energy consumption of the power transmission and distribution equipment.

[0046] Furthermore, in the system model concretization module, the pre-constructed system model includes a geostationary satellite GEO, N 5G base stations, and M power transmission and distribution equipment. The task segmentation model is used to generate A for each power transmission and distribution equipment m in each time slotm (t) The computing tasks to be processed by the unit are divided into tasks for local computing by the unit, tasks unloaded by the unit to the 5G base station, tasks unloaded by the unit to the geostationary satellite GEO, and the tasks for local computing, the tasks unloaded to the 5G base station, and the tasks unloaded to the geostationary satellite GEO are executed in parallel;

[0047] The power transmission and distribution equipment m is provided with buffer queues for storing the tasks for local computing, the tasks unloaded to the 5G base station, and the tasks unloaded to the geostationary satellite GEO respectively and

[0048] Furthermore, in the system model optimization problem establishment module, the system model optimization problem is specifically the long-term average energy consumption optimization problem of the power transmission and distribution equipment, expressed as:

[0049] P1:

[0050] s.t.C 1 :

[0051] C 2 :

[0052] C 3 :

[0053] C 4 :

[0054] C 5 :

[0055] C 6 :

[0056] Among them, the optimized variables are the task segmentation A(t) of the power transmission and distribution equipment, A(t) = {A m (t), m ∈ M}, the offloading decision a(t) of the task, and the computing resource allocation f n (t) of the 5G base station, C 1 represents the task allocation limit of each power transmission and distribution equipment, C 2 represents that the computing power allocated to the power transmission and distribution equipment is limited by the available computing resources of the 5G base station, C 3 and C 4 represent that each power transmission and distribution equipment can select at most one base station for connection in each time slot, C 5 ensures the delay constraint of the system, C 6It represents the non-negativity of the optimized variables. M represents the set of power transmission and distribution equipment, N represents the set of 5G base stations. When the total optimization time T = {1,..., t,... T} is divided into T time slots, an equal time slot model is adopted.

[0057] Furthermore, the optimal solution solving module is specifically configured to decouple the energy consumption optimization problem within a single time slot into three independent sub-problems, namely, the power transmission and distribution equipment task segmentation problem SP1, the power transmission and distribution equipment and 5G base station association control optimization problem SP2, and the 5G base station computing resource allocation problem SP3.

[0058] Compared with the prior art, the present invention has the following beneficial technical effects:

[0059] The present invention proposes a task offloading and resource optimization method for a 5G and satellite integrated network, aiming to minimize the long-term average energy consumption of power transmission and distribution equipment while considering the average delay of the system. The present invention decouples the long-term stochastic optimization problem into deterministic optimization problems for each time slot, and each deterministic optimization problem for a time slot is decomposed into three sub-problems. By sequentially solving the three sub-problems, the computing resource optimization of ground 5G base stations is completed. The present invention has very important scientific research significance and value for the resource management of the integration of 5G and satellite networks.

[0060] Furthermore, the present invention solves task segmentation through the Lagrangian dual method, obtains the best association between power transmission and distribution equipment and ground 5G base stations by using a many-to-one matching game, and proposes a resource allocation method based on minimizing the objective function with queue awareness, which can provide on-demand computing capabilities for power transmission and distribution equipment, minimize the long-term average energy consumption of power transmission and distribution equipment, and improve network performance. Description of the Drawings

[0061] The drawings in the specification are used to provide a further understanding of the present invention and constitute a part of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention.

[0062] Figure 1 It is a flowchart of the method according to an embodiment of the present invention.

[0063] Figure 2 It is a schematic diagram of the system model according to an embodiment of the present invention. Detailed Embodiments

[0064] To enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0065] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0066] Embodiment 1

[0067] A task offloading and resource optimization method for a 5G and satellite integrated network for the power transmission and distribution scenario, as Figure 1 shown, includes the following steps:

[0068] S1. Build a system model;

[0069] As Figure 2 shown, assuming in the considered power transmission and distribution scenario, it includes a geostationary satellite GEO, N 5G base stations and M power transmission and distribution devices. M and N are used to represent the set of power transmission and distribution devices and the set of 5G base stations respectively. Among them, the application programs running on the ground power transmission and distribution devices may generate computing tasks to be executed, and these devices often have very limited energy and computing capabilities. Therefore, in the model, task offloading and resource optimization are considered in the 5G and satellite integrated network. The geostationary satellite GEO and 5G base stations respectively provide computing capabilities for the power transmission and distribution devices, relieve the computing pressure of the power transmission and distribution devices, and at the same time reduce the energy consumption of the power transmission and distribution devices. In the model, the power transmission and distribution devices, 5G base stations and geostationary satellite GEO operate in different frequency bands, so they do not interfere with each other.

[0070] The total optimization time is divided into T time slots. The model adopts a discrete equal-time-slot model, with each time slot having a length of τ. The time-slot model is expressed as T = {1,..., t,... T}. Within each time slot, the state information of the entire network or system remains unchanged, and each power transmission and distribution device independently determines the splitting and offloading strategies for the tasks it generates. However, the state information of the system changes between different time slots.

[0071] S2. Specify the system model;

[0072] A. Task splitting model

[0073] In each time slot, each power transmission and distribution device m generates A m (t) (bits) of computing tasks to be processed. Specifically, we consider a task splitting model oriented towards data splitting, where the arriving tasks are split through an optimization algorithm and executed in parallel, i.e., local computing offloaded to the 5G base station offloaded to the geostationary satellite GEO Each power transmission and distribution device has three task buffer queues large enough and to store the tasks for local computing, offloading to the 5G base station, and offloading to the geostationary satellite GEO.

[0074] B. Data transmission model

[0075] Each power transmission and distribution device is equipped with two communication interfaces for communicating with the 5G base station and the geostationary satellite GEO respectively. When, at time slot t, the power transmission and distribution device m has tasks to be transmitted to the nth 5G base station for computing, according to Shannon's formula, the data transmission rate R m (t) can be obtained; when, at time slot t, the power transmission and distribution device m has tasks to be transmitted to the geostationary satellite GEO for computing, due to the long distance between the ground and the geostationary satellite GEO, a fixed data transmission rate is adopted.

[0076] C. Computing model

[0077] At time slot t, when tasks are selected for local computing, the local computing resources are used to directly compute the amount of task data that can be completed in the current time slot.

[0078] At time slot t, when tasks are selected to be offloaded to the nth 5G base station for computing, due to the large number of power transmission and distribution devices and the limited computing resources of the 5G base station, the queue backlog W m,n (t) after the tasks are transmitted to the nth 5G base station needs to be considered, and then the amount of task data completed is obtained by optimizing the computing resources allocated by the 5G base station for the power transmission and distribution device.

[0079] At time slot t, when a task chooses to be offloaded to the geostationary satellite GEO for computing, since the computing power of the geostationary satellite GEO is very strong, it can be assumed that the arriving data packets can be processed without any queuing delay. Therefore, the queue backlog for computing on the GEO satellite is not considered.

[0080] D. System Delay Constraint Model

[0081] According to Little's theorem, when the arrival rate of data traffic tasks is constant, the average delay is proportional to the task queue length. This model ensures the average system delay by restricting the average time length of the task queue.

[0082] E. Energy Consumption Model

[0083] Calculate the sum of the local computing energy consumption of each power transmission and distribution device, the energy consumption for task offloading through the power transmission and distribution device - 5G base station, and the energy consumption for task offloading through the power transmission and distribution device - geostationary satellite GEO.

[0084] S3. Establish an optimization problem for the system model to minimize the long-term average energy consumption of the power transmission and distribution device, and propose reasonable constraint conditions;

[0085] P1:

[0086] s.t.C 1 :

[0087] C 2 :

[0088] C 3 :

[0089] C 4 :

[0090] C 5 :

[0091] C 6 :

[0092] Among them, the optimized variables are the task splitting A(t) of the device, the task offloading decision a(t), and the computing resource allocation f n (t). C 1 represents the task allocation limit of each power transmission and distribution device. C 2 represents that the computing power allocated to the power transmission and distribution device is limited by the available computing resources of the 5G base station. C 3 and C 4It is stated that each power transmission and distribution device can select at most one base station for connection in each time slot. C 5 Ensure the delay constraint of the system. C 6 Indicate the non-negativity of the optimized variables. The model focuses on the scenario where 5G base stations serve a large number of power transmission and distribution devices simultaneously, rather than sequential service.

[0093] S4. Transform the system model optimization problem;

[0094] Based on Lyapunov optimization theory, transform the long-term energy consumption optimization problem into an energy consumption optimization problem within a single time slot.

[0095] S5. Decouple the energy consumption optimization problem within a single time slot into three separate sub-problems and solve them sequentially to obtain the optimal solution of problem P1. The three sub-problems are the power transmission and distribution device task segmentation problem SP1, the power transmission and distribution device and 5G base station association control optimization problem SP2, and the 5G base station computing resource allocation problem SP3.

[0096] A. Power transmission and distribution device task segmentation problem

[0097] The power transmission and distribution device task segmentation problem SP1 is a convex problem and is solved using the method of Lagrangian dual decomposition.

[0098] B. Power transmission and distribution device and 5G base station association control optimization problem

[0099] Minimize problem SP2 through many-to-one matching games. It includes:

[0100] The first step: Initialize the set of power transmission and distribution devices, the set of 5G base stations, and the preference list of each power transmission and distribution device and 5G base station.

[0101] In each time slot, each power transmission and distribution device preferentially selects the channel with the highest signal-to-noise ratio in the set of 5G base stations to obtain a high offloading rate. Therefore, the signal-to-noise ratio between the power transmission and distribution device and the 5G base station is defined as the preference function of power transmission and distribution.

[0102] To reduce the total energy consumption of the devices, 5G base station n preferentially selects the power transmission and distribution device that can achieve the minimum value of the objective function Γ(a m,n (t)) in the set of available power transmission and distribution devices. Therefore, the value of Γ(a m,n (t)) of the power transmission and distribution device is defined as the preference function of the 5G base station.

[0103] The second step: The matching process starts from the power transmission and distribution device and selects the most preferred 5G base station n in the priority list P m (t) of the power transmission and distribution device m.

[0104] Step 3: If there is a vacancy in the selected 5G base station n, directly add the matching pair (m, n) to the matching set. Otherwise, if there is no vacancy in the 5G base station n, compare the power transmission and distribution equipment m with all the power transmission and distribution equipment currently matched with the 5G base station n. If the power transmission and distribution equipment m is better than the worst-matched power transmission and distribution equipment m', then exchange the power transmission and distribution equipment m' and the power transmission and distribution equipment m.

[0105] Step 4: When there is no blocking pair or the preference lists of all participants (power transmission and distribution equipment and 5G base stations) are empty, the matching terminates, and a stable matching between the power transmission and distribution equipment and the 5G base stations at time slot t is obtained, that is, the SP2 optimal solution is obtained.

[0106] Among them, the blocking pair is defined as: for the current matching, if two different power transmission and distribution equipment are randomly selected for exchange matching, and the utility of one or more participants increases while the utility of other participants does not decrease, it is called a blocking pair.

[0107] C. 5G Base Station Computing Resource Allocation Problem

[0108] When solving the 5G base station computing resource allocation problem SP3, a 5G base station computing resource allocation strategy based on minimizing the objective function with queue awareness is proposed. First, in each 5G base station, give priority to the computing task queue with the longest queue to meet its computing resource requirements, and obtain a computing resource allocation scheme based on queue awareness; second, initialize the computing resources of each 5G base station, and in each 5G base station, give priority to the computing task queue that can achieve the minimum SP3 objective function value to meet its computing resource requirements, and obtain a computing resource allocation scheme based on minimizing the objective function; finally, according to the weight factors of the computing resource allocation of the two schemes, the optimal solution of SP3 is obtained, so as to minimize the long-term average energy consumption of the power transmission and distribution equipment.

[0109] Embodiment 2

[0110] The task offloading and resource optimization system for the 5G and satellite integrated network includes:

[0111] System model concretization module: used to concretize the pre-constructed system model into a task segmentation model, a data transmission model, a computing model, a system delay constraint model, and an energy consumption model;

[0112] System model optimization problem establishment module: used to establish a system model optimization problem by using the task segmentation model, the data transmission model, the computing model, the system delay constraint model, and the energy consumption model; among them, the total optimization time of the system model is divided into T equal time slots;

[0113] Optimization problem transformation module: used to transform the long-term average energy consumption optimization problem of the power transmission and distribution equipment into an energy consumption optimization problem within a single time slot;

[0114] Optimal solution solving module: used to decouple the energy consumption optimization problem within the single time slot into at least one separate sub-problem, and sequentially solve the sub-problems to obtain an optimal solution that can minimize the long-term average energy consumption of the power transmission and distribution equipment.

[0115] In the system model concretization module, the pre-constructed system model includes a geostationary satellite GEO, N 5G base stations, and M power transmission and distribution equipment. The task segmentation model is used to segment the A m (t) unit of computing tasks to be processed into tasks for unit local computing, tasks for unit offloading to the 5G base station, tasks for unit offloading to the geostationary satellite GEO, and the tasks for local computing, the tasks offloaded to the 5G base station, and the tasks offloaded to the geostationary satellite GEO are executed in parallel;

[0116] Power transmission and distribution equipment m is provided with buffer queues for respectively storing the tasks for local computing, the tasks offloaded to the 5G base station, and the tasks offloaded to the geostationary satellite GEO and

[0117] In the system model optimization problem establishment module, the system model optimization problem is specifically the long-term average energy consumption optimization problem of the power transmission and distribution equipment, expressed as:

[0118] P1:

[0119] s.t.C 1 :

[0120] C 2 :

[0121] C 3 :

[0122] C 4 :

[0123] C 5 :

[0124] C 6 :

[0125] Among them, the optimized variable is the task segmentation A(t) of the power transmission and distribution equipment, A(t) = {A m(t), m ∈ M}, the offloading decision a(t) of the task, and the computing resource allocation f of the 5G base station n (t), C 1 represents the task allocation limit of each power transmission and distribution device, C 2 represents that the computing power allocated to the power transmission and distribution device is limited by the computing resources available at the 5G base station, C 3 and C 4 represents that each power transmission and distribution device can select at most one base station for connection in each time slot, C 5 ensures the delay constraint of the system, C 6 represents the non-negativity of the optimized variables. M represents the set of power transmission and distribution devices, N represents the set of 5G base stations, and when the total optimization time T = {1,..., t,... T} is divided into T time slots, an equal time slot model is adopted.

[0126] The optimal solution solving module is specifically configured to decouple the energy consumption optimization problem within a single time slot into three independent sub-problems, namely, the power transmission and distribution device task segmentation problem SP1, the power transmission and distribution device and 5G base station association control optimization problem SP2, and the 5G base station computing resource allocation problem SP3.

[0127] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0128] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, so that the instructions executed by the processors of the computer or other programmable data processing devices generate means for realizing the functions specified in one Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0129] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more of the processes Figure 1 one or more processes and / or blocks Figure 1 specified in the block or blocks.

[0130] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more of the processes Figure 1 one or more processes and / or blocks Figure 1 specified in the block or blocks.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the scope of its protection. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: after reading the present invention, those skilled in the art can still make various changes, modifications, or equivalent replacements to the specific embodiments of the invention, but these changes, modifications, or equivalent replacements are all within the scope of the claims of the invention pending approval.

Claims

Task offloading and resource optimization method for 1.5G and satellite integrated network Characterized in that Including Specify the pre - constructed system model into a task segmentation model, a data transmission model, a computing model, a system delay constraint model, and an energy consumption model Among them, the pre-constructed system model includes a geostationary satellite GEO, N 5G base stations, and M power transmission and distribution devices. The task segmentation model is used to divide the A m (t) unit of computing tasks to be processed into tasks for local computing per unit, tasks for offloading to 5G base stations per unit, tasks for offloading to the geostationary satellite GEO per unit, and the tasks for local computing, the tasks for offloading to 5G base stations, and the tasks for offloading to the geostationary satellite GEO are executed in parallel; The power transmission and distribution equipment m is provided with buffer queues for storing tasks calculated locally, tasks unloaded to the 5G base station, and tasks unloaded to the geostationary satellite GEO respectively and The data transmission model is used to calculate the data transmission rate. Specifically, each power transmission and distribution device is equipped with two communication interfaces, which communicate with a 5G base station and a geostationary satellite GEO respectively When the power transmission and distribution equipment m has a task to be transmitted to the nth 5G base station for calculation at time slot t, the data transmission rate R is obtained according to the Shannon formula m (t); when the power transmission and distribution equipment m has a task to be transmitted to the geostationary satellite GEO for calculation at time slot t, a fixed data transmission rate is adopted; The computing model is used to calculate the amount of task data completed in the current time slot. Specifically At time slot t, when a task chooses to be computed locally, use local computing resources to calculate the amount of task data that can be completed in the current time slot At time slot t, when a task is selected to be offloaded to the nth 5G base station for computing, consider the queue backlog W m,n (t) after the task is transmitted to the nth 5G base station, and then obtain the amount of task data completed by optimizing the computing resources allocated by the 5G base station to the power transmission and distribution equipment; At time slot t, when a task chooses to be offloaded to the geostationary satellite GEO for computing, multiply the time slot length by the data transmission rate from the ground to the satellite to obtain the amount of task data completed The system delay constraint model ensures the average system delay by restricting the average time length of the task queue. The task queue includes the buffer queue of tasks and the backlog queue of tasks transmitted to the 5G base station The energy consumption model is used to calculate the sum of the local computing energy consumption of each power transmission and distribution device, the energy consumption of task offloading from the power transmission and distribution device to the 5G base station, and the energy consumption of task offloading from the power transmission and distribution device to the geostationary satellite GEO Establish a system model optimization problem using the task segmentation model, data transmission model, computing model, system delay constraint model, and energy consumption model. Among them, the total optimization time of the system model is divided into T equal time slots The system model optimization problem is specifically the long - term average energy consumption optimization problem of the power transmission and distribution device, expressed as Among them, the optimized variables are the task segmentation A(t) of the power transmission and distribution equipment, A(t) = {A m (t), m ∈ M}, the task offloading decision a(t), and the computing resource allocation f n (t), C 1 represents the task allocation limit of each power transmission and distribution equipment, C 2 represents that the computing power allocated to the power transmission and distribution equipment is limited by the computing resources available at the 5G base station, C 3 and C 4 represent that each power transmission and distribution equipment can select at most one base station for connection in each time slot, C 5 ensures the delay constraint of the system, C 6 represents the non-negativity of the optimized variables, M represents the set of power transmission and distribution equipment, N represents the set of 5G base stations, and when the total optimization time T = {1,..., t,... T} is divided into T time slots, the equal time slot model is adopted; Convert the long - term average energy consumption optimization problem of the power transmission and distribution device into an energy consumption optimization problem within a single time slot Decouple the energy consumption optimization problem within the single time slot into at least one separate sub - problem, and solve the sub - problems in sequence to obtain the optimal solution that can minimize the long - term average energy consumption of the power transmission and distribution device 2. The task offloading and resource optimization method for 5G and satellite integrated network according to claim 1 Characterized in that Decouple the energy consumption optimization problem within the single time slot into at least one separate sub - problem. Specifically, decouple the energy consumption optimization problem within the single time slot into three independent sub - problems, namely the power transmission and distribution device task segmentation problem SP1, the power transmission and distribution device and 5G base station association control optimization problem SP2, and the 5G base station computing resource allocation problem SP3 3. The task offloading and resource optimization method for 5G and satellite integrated network according to claim 1 Characterized in that The power transmission and distribution device and 5G base station association control optimization problem SP2 is realized through a many - to - one matching game, specifically including Initialize the set of power transmission and distribution devices, the set of 5G base stations, and the preference lists of each power transmission and distribution device and 5G base station The matching process starts from the power transmission and distribution equipment, and selects the most preferred 5G base station n from the priority list P m (t) of the power transmission and distribution equipment m; If there is available capacity in the selected 5G base station n, the matching pair (m, n) is added to the matching set. If there is no available capacity in the 5G base station n, the power transmission and distribution equipment m is compared with all the power transmission and distribution equipment currently matched with the 5G base station n. If the power transmission and distribution equipment m is better than the worst-matched power transmission and distribution equipment m', then the power transmission and distribution equipment m' and the power transmission and distribution equipment m are exchanged; When there is no blocking pair or the preference lists of all power transmission and distribution equipment and 5G base stations are empty, the matching terminates, and a stable matching of power transmission and distribution equipment and 5G base stations at time slot t is obtained, thus obtaining the optimal solution of the power transmission and distribution equipment and 5G base station association control optimization problem SP2.

4. The task offloading and resource optimization method for the 5G and satellite integrated network according to claim 1, characterized in that, for solving the 5G base station computing resource allocation problem SP3, a 5G base station computing resource allocation strategy based on minimizing the objective function with queue awareness is adopted. The specific process is as follows: In each 5G base station, the computing task queue with the longest queue is given priority, and the computing resource requirements of the computing task queue with the longest queue in each 5G base station are satisfied to obtain a computing resource allocation scheme based on queue awareness; Initialize the computing resources of each 5G base station. In each 5G base station, the computing task queue that can achieve the minimum SP3 objective function value is given priority, and its computing resource requirements are satisfied to obtain a computing resource allocation scheme based on minimizing the objective function; According to the weight factors of the computing resource allocation of the above two computing resource allocation schemes, the optimal solution of the 5G base station computing resource allocation problem SP3 is obtained, thereby minimizing the long-term average energy consumption of the power transmission and distribution equipment.

5. The task offloading and resource optimization system for the 5G and satellite integrated network, characterized in that, including: System model concretization module: used to concretize the pre-constructed system model into a task segmentation model, a data transmission model, a computing model, a system delay constraint model, and an energy consumption model; Among them, the pre-constructed system model includes a geostationary satellite GEO, N 5G base stations, and M power transmission and distribution devices. The task segmentation model is used to divide the A m (t) unit of computing tasks to be processed into tasks for local computing of the unit, tasks for offloading to 5G base stations of the unit, tasks for offloading to the geostationary satellite GEO of the unit, and the tasks for local computing, the tasks for offloading to 5G base stations, and the tasks for offloading to the geostationary satellite GEO are executed in parallel; The power transmission and distribution equipment m is provided with buffer queues for storing tasks calculated locally, tasks unloaded to the 5G base station, and tasks unloaded to the geostationary satellite GEO, respectively. and The data transmission model is used to calculate the data transmission rate. Specifically: each power transmission and distribution equipment is equipped with two communication interfaces, which communicate with the 5G base station and the geostationary satellite GEO respectively; When the power transmission and distribution equipment m has a task to be transmitted to the nth 5G base station for calculation at time slot t, the data transmission rate R is obtained according to the Shannon formula m (t); when the power transmission and distribution equipment m has a task to be transmitted to the geostationary satellite GEO for calculation at time slot t, a fixed data transmission rate is adopted. The computing model is used to calculate the amount of task data completed in the current time slot. Specifically: At a time slot t When a task is selected for local computing, the local computing resources are used to calculate the amount of task data that can be completed in the current time slot; At time slot t, when a task is selected to be offloaded to the nth 5G base station for computing, consider the queue backlog W m,n (t) after the task is transmitted to the nth 5G base station, and then obtain the amount of task data completed by optimizing the computing resources allocated by the 5G base station to the power transmission and distribution equipment; At time slot t, when a task is selected to be offloaded to the geostationary satellite GEO for computing, the amount of task data completed is obtained by multiplying the time slot length by the data transmission rate from the ground to the satellite; The system delay constraint model ensures the average delay of the system by restricting the average time length of the task queue. The task queue includes the buffer queue of the task and the backlog queue of the task transmitted to the 5G base station; The energy consumption model is used to calculate the sum of the local computing energy consumption of each power transmission and distribution equipment, the energy consumption of task offloading from the power transmission and distribution equipment to the 5G base station, and the energy consumption of task offloading from the power transmission and distribution equipment to the geostationary satellite GEO; System model optimization problem establishment module: used to establish a system model optimization problem by using the task segmentation model, the data transmission model, the computing model, the system delay constraint model, and the energy consumption model; wherein, the total optimization time of the system model is divided into T equal time slots; The optimization problem of the system model is specifically the long-term average energy consumption optimization problem of power transmission and distribution equipment, which is expressed as: Among them, the optimized variables are the task segmentation A(t) of the power transmission and distribution equipment, A(t) = {A m (t), m ∈ M}, the task offloading decision a(t), and the computing resource allocation f n (t), C 1 represents the task allocation limit of each power transmission and distribution equipment, C 2 represents that the computing power allocated to the power transmission and distribution equipment is limited by the computing resources available at the 5G base station, C 3 and C 4 represent that each power transmission and distribution equipment can select at most one base station for connection in each time slot, C 5 ensures the delay constraint of the system, C 6 represents the non-negativity of the optimized variables, M represents the set of power transmission and distribution equipment, N represents the set of 5G base stations, and when the total optimization time T = {1,..., t,... T} is divided into T time slots, the adopted equal time slot model; Optimization problem transformation module: used to transform the long-term average energy consumption optimization problem of power transmission and distribution equipment into an energy consumption optimization problem within a single time slot; Optimal solution solving module: used to decouple the energy consumption optimization problem within the single time slot into at least one separate sub-problem, and sequentially solve the sub-problems to obtain an optimal solution that can minimize the long-term average energy consumption of power transmission and distribution equipment.

6. The task offloading and resource optimization system for the 5G and satellite integrated network according to claim 5, characterized in that the optimal solution solving module is specifically used to decouple the energy consumption optimization problem within the single time slot into three independent sub-problems, namely the task segmentation problem SP1 of power transmission and distribution equipment, the associated control optimization problem SP2 of power transmission and distribution equipment and 5G base stations, and the computing resource allocation problem SP3 of 5G base stations.

Images