Space-air-ground integrated network resource management method and system and storage medium

Abstract

The invention provides a space-air-ground integrated network resource management method and system and a storage medium, and belongs to the technical field of wireless communication and network resource dispatch. The method comprises the steps that a space-air-ground integrated network is constructed, a GU is provided with an energy collection module and a backscatter circuit, and the backscatter circuit is connected with the GU; low-power-consumption data transmission and energy collection are realized by reflecting a radio frequency carrier signal emitted by the UAV; the UAV carries a mobile edge computing server to process a task, and the task is unloaded to the LEO when the task is overloaded; a central controller collects network state and channel state information in real time, a random optimization function is constructed by taking minimization of long-term average energy consumption of a system as a target and combining constraints such as task queue stability and energy sustainability, a Lyapunov optimization framework is adopted to convert the problem into a single time slot problem, and the problem is solved through alternate optimization and concave-convex process technologies. According to the method, the stability of the task queue is guaranteed, the system energy consumption is remarkably reduced, the resource utilization rate is improved, and reliable support is provided for remote area power emergency communication.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication and network resource scheduling technology, specifically relating to an integrated air-space-ground network resource management method, system, and storage medium. Background Technology

[0002] In remote areas, communication infrastructure is weak or absent. When faced with sudden accidents (such as heavy rain and snow), traditional terrestrial networks have poor coverage and insufficient emergency response capabilities, which cannot meet the real-time data transmission needs of power facilities.

[0003] With the rapid growth in demand for Power Internet of Things (PIoT) services, an increasing number of power facilities are being deployed in remote areas with weak or absent communication infrastructure. The SAGIN integrated space-air-ground network, as a key potential architecture for new power systems, offers enhanced service capabilities in terms of coverage and flexibility.

[0004] The ground segment of the SAGIN integrated air-space-ground network consists of numerous PIoT ground user GUs, whose computing power and energy are limited. Furthermore, due to size and cost constraints, power IoT devices typically rely on embedded batteries with limited energy, making it difficult to guarantee continuous battery power to maintain network continuity during emergencies such as grid outages and emergency fault repairs. SAGIN employs a binary decision-making process of "fully local computing" or "fully offloaded computing," which cannot flexibly allocate tasks based on task urgency and device battery power, failing to meet the fine-grained computing needs of emergency scenarios. Traditional ground-based backscatter networks suffer severe degradation under non-line-of-sight (NLoS) conditions, with signal attenuation significantly reducing communication range and scalability. This results in a lack of dynamic coordinated scheduling of communication, computing, and energy resources across the air, space, and ground layers, easily leading to an imbalance of tasks backlogged on ground devices. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides a method, system, and storage medium for integrated air-space-ground network resource management.

[0006] To achieve the above objectives, the present invention provides a method for integrated air-space-ground network resource management, comprising: Construct an integrated air-space-ground network for the target power system, comprising a ground layer, an air layer, and a space layer; the ground layer consists of multiple ground-based devices (GUs), the air layer deploys unmanned aerial vehicles (UAVs), and the space layer deploys a single low-Earth orbit (LEO) satellite to handle offloading computational tasks assigned by the UAVs.

[0007] Each GU continuously collects real-time status data of the target power system according to the first task queue format; the UAV actively transmits radio frequency carrier signals to the GU at a preset frequency. After receiving the radio frequency carrier signals, the GU converts the radio frequency carrier signals into electrical energy storage, modulates the real-time status data onto the radio frequency carrier signals, and backscatters the modulated real-time status data back to the corresponding UAV to complete the task offloading; the UAV decodes the real-time status data modulated by the corresponding GU according to the second task queue format, performs calculation tasks, and transmits the calculation results back to the corresponding GU.

[0008] The central controller collects network status information of GU and UAV, air-to-ground channel status of UAV and GU, and air-to-space channel status of UAV and LEO in real time. Based on all the collected information, with the goal of minimizing the average energy consumption of network operation and with the stability of the two task queues and the stability of GU power as constraints, a stochastic optimization function is established for the time allocation, task offloading allocation and computing resource allocation of the network. The stochastic optimization function is solved using the Lyapunov drift penalty function method to obtain the optimal resource allocation scheme among GU, UAV and LEO.

[0009] Preferably, the real-time status data includes system fault information, system equipment monitoring data, and sensing data of the system's environment; the network status information of the GU includes the device's remaining energy, the length of the first task queue, and the number of new tasks arriving; the network status information of the UAV includes the load status of the second task queue, the computing resource occupancy, and its own remaining energy status; the air-to-ground channel status corresponds to the link attributes between the UAV and the GU, including channel gain, signal attenuation, and interference level; the air-space channel status corresponds to the link attributes between the UAV and the LEO, including link bandwidth, propagation delay, and signal stability; the GU's power stability specifically means that the power collected by the GU is greater than the power it consumes.

[0010] Preferably, the constraints further include: network transmission rate constraints, reflection coefficient constraints, computational resource constraints for GU and UAV, and air-to-space link delay constraints. The network transmission rate includes the air-to-ground link transmission rate between UAV and GU and the air-to-space link transmission rate between UAV and LEO, and the transmission rate must meet the data transmission requirements of the corresponding link. The reflection coefficient is the signal reflection parameter when GU modulates real-time state data, and its value range is [0,1]. The computational resource constraints for GU and UAV include: the local computation of GU does not exceed its own computational resource limit, and the local computation of UAV does not exceed the computational resource limit of its mobile edge computing (MEC) server. The air-to-space link delay constraint means that the sum of the transmission delay of UAV offloading tasks to LEO and the task processing delay of LEO does not exceed a preset threshold.

[0011] Preferably, the step of solving the stochastic optimization function using the Lyapunov drift penalty function method to obtain the optimal resource allocation scheme among GU, UAV, and LEO specifically includes: using the Lyapunov drift penalty function method, by introducing control parameters... V The long-term stochasticity of the stochastic optimization function is transformed into a deterministic optimization problem that depends only on the current time slot state, and then solved using an alternating optimization algorithm. The non-convex part of the air-to-ground link transmission rate with respect to the reflection coefficient is linearized using the concave-convex process CCCP technique, and the reflection coefficient, task offloading ratio, and computational resource allocation are solved. Based on the solution, the time allocation for energy harvesting, backscatter communication, task processing, and data transmission is further optimized, ultimately obtaining the optimal resource allocation scheme among the GU, UAV, and LEO, which is then distributed to each device in the GU, UAV, and LEO for execution. The control parameters... V Used to balance the average energy consumption of network operation with the stability of task queues.

[0012] Preferably, the GU is equipped with an energy harvesting module and a backscatter circuit. The energy harvesting module converts the received radio frequency carrier signal into electrical energy for storage, and the backscatter circuit modulates real-time status data onto the received radio frequency carrier signal. The UAV is equipped with a carrier generator and a mobile edge computing (MEC) server. The carrier generator transmits radio frequency carrier signals, and the MEC server performs local computing tasks.

[0013] Preferably, the local computation of the GU includes: dynamically adjusting the computation frequency through dynamic voltage and frequency scaling (DVFS) technology, and determining the amount of locally processed data within time slot t based on its own computational resource limit.

[0014] Preferably, the task offloading allocation includes the ratio of local computation of the GU to tasks offloaded to the corresponding UAV, and the ratio of local computation of the UAV to tasks offloaded to the corresponding LEO. The computation tasks adopt a partial offloading method, which is divided into a local execution part, an offloaded execution part to the UAV, and an offloaded execution part to the LEO, and the three are executed in parallel.

[0015] This invention also provides an integrated air-space-ground network resource management system, comprising: The construction module is used to build an integrated air-space-ground network for the target power system. The network includes a ground layer, an air layer, and a space layer. The ground layer consists of multiple ground devices (GUs), the air layer is equipped with unmanned aerial vehicles (UAVs), and the space layer is equipped with a single low-Earth orbit (LEO) satellite to undertake offloading computing tasks assigned by the UAVs.

[0016] The computing module is used for each GU to continuously collect real-time status data of the target power system according to the first task queue. The UAV actively transmits radio frequency carrier signals to the GU at a preset frequency. After receiving the radio frequency carrier signals, the GU converts the radio frequency carrier signals into electrical energy storage, modulates the real-time status data onto the radio frequency carrier signals, and backscatters the modulated real-time status data back to the corresponding UAV to complete the task offloading. The UAV decodes the real-time status data modulated by the corresponding GU according to the second task queue, performs the computing task, and transmits the computing results back to the corresponding GU.

[0017] The allocation module is used by the central controller to collect network status information of GU and UAV, air-to-ground channel status of UAV and GU, and air-to-space channel status of UAV and LEO in real time. Based on all the collected information, with the goal of minimizing the average energy consumption of network operation and with the stability of the two task queues and the stability of GU power as constraints, a stochastic optimization function is established for the time allocation, task offloading allocation and computing resource allocation of the network. The stochastic optimization function is solved using the Lyapunov drift penalty function method to obtain the optimal resource allocation scheme among GU, UAV and LEO.

[0018] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement any of the steps in the integrated air-space-ground network resource management method.

[0019] The present invention also provides a computer-readable storage medium storing a computer program, which, when loaded by a processor, is capable of executing any of the steps in the integrated air-space-ground network resource management method.

[0020] The integrated air-space-ground network resource management method provided by this invention has the following beneficial effects: The ground-level GU of this invention uses backscatter communication to modulate real-time power system status data collected from the UAV onto a carrier wave and reflect it back, simultaneously completing energy harvesting. This overcomes the non-line-of-sight limitations of traditional ground-based backscatter communication by utilizing the UAV's controllable line-of-sight link, and achieves extremely low-power uplink communication and energy replenishment from the source. The air-level UAV receives information returned by the GU and processes tasks, while the space-level LEO serves as a backup computing node. Based on this, the invention collects real-time network status information and constructs a stochastic optimization function with the objective of minimizing long-term average energy consumption, constrained by the stability of the dual queues of GU and UAV and the causality of GU energy. This function is solved online using a Lyapunov optimization framework, dynamically outputting the optimal resource allocation scheme for each time slot. This scheme jointly optimizes time allocation, the fine-grained offloading ratio of tasks on the GU, between the UAV and LEO, and the computing resource allocation of each node.

[0021] This invention solves the energy bottleneck problem through low-power backscatter communication, ensures communication reliability through UAV-assisted line-of-sight links, and completely abandons the rigid binary offloading mode through online dynamic joint resource scheduling based on Lyapunov optimization. It realizes adaptive coordination of air, space, and ground communication, computing, and energy resources, and ultimately achieves the minimization of overall system energy consumption and sustainable operation while ensuring that the task queue does not back up. Attached Figure Description

[0022] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating an integrated air-space-ground network resource management method according to an embodiment of the present invention. Figure 2 This is a diagram of the backscattering-assisted SAGIN architecture model according to an embodiment of the present invention. Figure 3 This diagram illustrates the network system stability performance according to an embodiment of the present invention. Figure 4 Control parameters for embodiments of the present invention V Impact diagram on network system performance Figure 4 (a) is a graph showing the impact on queue backlog. Figure 4 (b) is a diagram showing the impact on energy consumption; Figure 5 The task arrival rate in this embodiment of the invention Impact on network system performance; Figure 5 (a) is a graph showing the impact on queue backlog. Figure 5 (b) is a graph showing the impact on energy consumption. Detailed Implementation

[0024] To enable those skilled in the art to better understand and implement the technical solutions of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0025] This invention proposes a backscatter-assisted SAGIN architecture specifically tailored for emergency communications. In the constructed hierarchical SAGIN system, the UAV uses backscatter communication (BackCom) to assist the ground-based GU in task offloading and decision-making. Tasks are executed on the UAV and further offloaded to the LEO satellite to support stable data processing in PIoT during emergencies. Both the UAV and the LEO satellite nodes are equipped with mobile edge computing (MEC) servers. Under constraints of network queue stability and energy causality, a long-term average energy consumption minimization problem is proposed by jointly optimizing time allocation, reflection coefficient offloading strategies, and computational resources. The task queue is dynamically modeled, and the stochastic long-term optimization is transformed into a tractable single-slot problem using a Lyapunov optimization framework. Due to the inherent coupling between variables, the problem is decomposed into two subproblems, solved by introducing slack variables and the concave-convex process CCCP. Simulation results verify that, compared with benchmark schemes, the proposed algorithm achieves a near-optimal trade-off between meeting energy-saving requirements and maintaining queue stability.

[0026] Based on this, this invention provides an integrated air-space-ground network resource management method for power emergency communication scenarios, specifically proposing an integrated air-space-ground network resource management method that integrates low-power backscatter communication. The core of this method lies in using a dynamic collaborative control mechanism to intelligently command ground equipment, drones, and low-orbit satellites, minimizing the long-term energy consumption of the entire system while ensuring that computing tasks do not backlog. Specifically, as follows... Figure 1 As shown, it includes:

[0027] S1. Construct an integrated air-space-ground network for the target power system, comprising a ground layer, an air layer, and a space layer; the ground layer consists of multiple ground-based devices (GUs), the air layer deploys unmanned aerial vehicles (UAVs), and the space layer deploys a single low-Earth orbit (LEO) satellite to handle offloading computational tasks assigned by the UAVs.

[0028] S101, Backscatter-assisted SAGIN architecture and collaborative workflow like Figure 2 As shown, the proposed backscatter-assisted SAGIN structure is a three-layer architecture, consisting of a ground layer, an aerial layer, and a space layer. The ground layer is composed of... Composed of GU units, using This indicates that each GU is equipped with an energy harvesting (EH) module and a backscatter circuit, enabling low-power passive communication via external radio frequency (RF) signals in emergency communication scenarios. The air layer consists of single UAVs. As the mobile RF power source for the GU, it assists in handling complex computing tasks and is equipped with a carrier generator and MEC server. The space layer includes a low-Earth orbit satellite, denoted as... Used as a remote computing node, the satellite's communication range can cover the UAV and multiple ground equipment GUs in the entire emergency communication scenario.

[0029] S2. Each GU continuously collects real-time status data of the target power system according to the first task queue format; the UAV actively transmits radio frequency carrier signals to the GU at a preset frequency. After receiving the radio frequency carrier signals, the GU converts the radio frequency carrier signals into electrical energy storage, and simultaneously modulates the real-time status data onto the radio frequency carrier signals, and backscatters the modulated real-time status data back to the corresponding UAV to complete task offloading; the UAV decodes the real-time status data modulated by the corresponding GU according to the second task queue format, performs calculation tasks, and transmits the calculation results back to the corresponding GU.

[0030] Divide the total execution time of the optimization process evenly into equal parts. A discrete interval, using This indicates that each time slot... constant length For ease of analysis, a three-dimensional Cartesian coordinate system is used, where GU has a fixed position. Considering the maneuverability of UAVs, the UAVs maintain a fixed flight altitude. Its time-varying position is from Given. To ensure continuous operation, the UAV's motion trajectory is specified as a circle centered on the simulation area, with a radius of... They move in circular motion. The motion of a low-Earth orbit satellite is determined by its orbital parameters. ,in This represents the orbital radius from the Earth's center. It is the track inclination angle. It is the longitude of the ascending node. This indicates that the satellite and the rising intersection point are in the time slot. The included angle.

[0031] This invention divides each time slot of the proposed network into four different stages, namely: Energy harvesting phase The UAV transmits radio frequency signals to the GU. All GUs switch to energy harvesting mode, converting the received radio frequency signals into electrical energy for storage.

[0032] Backscattering stage The GU switches to BackCom mode. Following instructions from the control center, the GU modulates the computational task data to be unloaded onto the incident RF signal emitted by the UAV using a specific reflection coefficient, and then reflects it back to the UAV.

[0033] Relay forwarding The UAV assesses the load on its own task queue. If the load is too heavy, some tasks received from the GU will be proactively uploaded to the more powerful LEO for processing via the space-to-space link.

[0034] Result feedback stage The edge computing server sends the task processing results to the corresponding GU. Since the result data volume is small and the downlink rate is high, the latency at this stage is negligible.

[0035] S201, System Model Concretization Air-to-Ground Link: Based on Shannon's theorem, the maximum transmission rate of the air-to-ground passive link is obtained. The signal decoding process utilizes continuous interference cancellation (SIC) to decode the superimposed signals one by one. The decoding order is determined by the descending order of the channel gain, denoted as... ,in Corresponding to the channel gain of the first Strongest user equipment. Typically, the strongest GU is decoded first, as it withstands interference from all other users, while the weakest GU can be decoded without interference.

[0036] Air-to-space link: During the relay phase, the UAV actively relays the computing task to low Earth orbit, from which the achievable uplink rate of the air-to-space link can be obtained. .

[0037] Assuming the computational workload of GU Random arrival slot This satisfies the Independent and Identically Distributed (iID) process. In partial offloading, the computational task is divided into two or more independent parts, which are executed in parallel through local computation and task offloading. This is more suitable for the emergency communication scenario of damaged ground communication networks considered in this invention. In time slots The workloads for local execution, unloading to UAV, and unloading to LEO are respectively , and .

[0038] The GPU's computing frequency is dynamically controlled through Dynamic Voltage and Frequency Scaling (DVFS) technology. Therefore, based on the GPU's own computing power, time slots can be obtained. Local data processing volume. In emergency situations where GU power is severely limited, user task demands cannot be met. To alleviate this situation, BackCom allows some tasks to be offloaded to UAVs for execution, thereby increasing the data processing volume of UAV time slots t. Compared to the task queuing method used for UAV offloading, LEOs are typically equipped with multi-core CPUs, enabling parallel task computation, and queuing latency is negligible. Furthermore, unlike air-to-ground transmission, the propagation latency is very small due to the short distance. Offloading tasks to LEOs would introduce significant propagation latency, creating a propagation latency constraint.

[0039] Due to signal attenuation in the air-to-ground link, the incident power of the GU is insufficient to activate the nonlinear region. Therefore, this invention employs a linear EH model to obtain the energy collected within time slot t. In backscattering... During this period, the GU modulates its offloaded tasks onto the RF signal without generating a carrier itself, thus only consuming circuit power. In addition, performing local computation generates computation-related energy consumption. Offload energy consumption consists of two parts: computation energy consumption and air-to-space uplink transmission energy consumption from the UAV to the LEO.

[0040] S202, Dynamic Resource Management and Optimization Control Methods To achieve long-term energy saving and queue stability, this invention proposes a dynamic optimization control process where each time slot is controlled by a central controller, executing the following steps: Information Sensing and Queue Update: The controller collects network status information for the current time slot, including: remaining power of the GUs, task buffer queue length, number of newly arrived computational tasks, air-to-ground channel quality, and air-to-space channel quality. Subsequently, based on the task processing status of the previous time slot, each GU's own task queue is updated using the following formula. Task queues between UAVs and each GU .

[0041] ; ; Among them, the function Ensure the backlog is non-negative.

[0042] ; Where C1 represents the time allocation constraint; C2 and C3 represent the offloading task amount constraint and the air-to-ground transmission rate constraint, respectively; and C4 represents the air-to-space link delay constraint. This is the air-to-ground distance. The speed of light is constant; C5 ensures that GU is in the time slot. Collect energy Exceeding its local energy consumption This ensures the sustainability of the network; C6 and C7 represent network stability constraints; C8 constrains the reflection coefficient of the GU; C9 and C10 specify the computational resource constraints for the GU and UAV, respectively. and These represent the maximum computing resources available for the GU and UAV, respectively.

[0043] As shown above, construct an immediate optimization problem. First, construct a method to jointly optimize time allocation. ), reflectance coefficient Task unloading strategy Local and UAV computing resources To minimize long-term average energy consumption The original problem .

[0044] S3. The central controller collects network status information of GU and UAV, air-to-ground channel status of UAV and GU, and air-to-space channel status of UAV and LEO in real time. Based on all the collected information, with the goal of minimizing the average energy consumption of network operation and with the stability of the two task queues and the stability of GU power as constraints, a stochastic optimization function is established for the time allocation, task offloading allocation and computing resource allocation of the network. The stochastic optimization function is solved using the Lyapunov drift penalty function method to obtain the optimal resource allocation scheme among GU, UAV and LEO.

[0045] Next, based on the network state information collected in the current time slot, the complex stochastic optimization problem of minimizing long-term average energy consumption is transformed into a deterministic optimization problem that depends only on the state of the current time slot. This transformation is achieved through the Lyapunov optimization framework, the core of which is the introduction of a control parameter. It is used to flexibly balance the two goals of "reducing energy consumption" and "clearing the task queue".

[0046] ; Alternating optimization is used to solve the decision variables, where the immediate optimization problem requires simultaneous decisions on multiple variables. Because these variables are coupled and the problem is non-convex, this invention employs an efficient alternating iterative algorithm for solution.

[0047] Subproblem 1: Reflection Coefficient and Resource Optimization. Allocating resources within a given time frame... Under this scheme, the backscattering coefficients, task offloading ratios, and computational resource allocation of each GU are optimized. .

[0048] ; In this problem, the air-to-ground transmission rate constraint C3 is relative to the reflection coefficient. It is non-convex. Therefore, this invention employs CCCP technology, in which... For the convex part, linearize the non-convex part to obtain the following expression:

[0049] ; in, It is the available transmission bandwidth. It's GU The reflection coefficient, Indicates the UAV's transmission power. This is the air-to-ground channel gain. This represents interference from other GUs. This represents the power of additive white Gaussian noise (AWGN).

[0050] Subsequently, sub-problems It is equivalently transformed into the following expression: ; At this point, the subproblem It has been transformed into a convex optimization problem, which can be effectively solved using the CVX toolbox.

[0051] Sub-problem 2: Time coefficient optimization. Given parameters such as reflection coefficient and unloading ratio... Then, optimize the time allocation for the four stages. .

[0052] ; At this point, the problem is transformed into a convex optimization problem, which can be solved directly.

[0053] Command issuance and execution: The controller issues and executes the optimized decision. Each layer of the network executes communication, computation, and energy harvesting operations in its current time slot according to the command. Based on the above analysis, the specific dynamic optimization control flow is shown in Table 1.

[0054] Table 1 Dynamically Optimized Control Flow This invention studies the impact of different simulation parameters on algorithm performance and compares the results with three benchmark schemes in terms of system queue backlog and energy consumption to evaluate the performance of the proposed algorithm.

[0055] Simulation Setup. This invention considers an emergency communication scenario, including a LEO satellite, a UAV, and multiple disaster-affected GUs randomly distributed within a 100m × 100m area. The UAV flies at a fixed altitude and follows a fixed trajectory within the simulation area, while the LEO satellite covers the entire region. All parameters not specifically mentioned use the default values ​​in Table 2.

[0056] Table 2 Simulation Parameter Settings To highlight the designed algorithm, this invention compares the proposed method with four benchmark schemes. The Energy Minimization Method (EMA) calculates the minimum system energy consumption without considering long-term queue stability. In the NOMA-based Full Offloading Method (NCOA), the GU (Guard Unit) reuses the air-to-ground channel via NOMA to offload tasks. In the OMA-based Partial Offloading Method (OPOA), the GU uses OMA, resulting in even bandwidth distribution and no interference. The Local Computation Method (LCA) processes computation tasks locally, satisfying the long-term stability constraint C6 of the local queue.

[0057] Simulation results analysis. Figure 3 The system performance over time is shown for different algorithms. Compared with EMA and LCA algorithms, NCOA, OPOA, and the proposed algorithm exhibit stable upper and lower bounds. The queue backlog does not increase continuously over time, demonstrating the importance of the Lyapunov optimization framework in maintaining data queue stability. However, because EMA focuses solely on energy saving, and LCA, which relies entirely on local computation, proves insufficient to manage the continuous arrival of tasks, queue backlog remains persistently high, making its growth uncontrollable. Furthermore, OPOA is constrained by network bandwidth and power limitations, leading to some underutilization of edge resources, while NCOA utilizes shared bandwidth but simplifies local computation and energy harvesting. Based on the above comparison, it is clear that the algorithm of this invention achieves the best trade-off and complementary performance in maintaining queue stability while meeting energy saving requirements.

[0058] Figure 4 The Lyapunov control parameters are displayed. Impact on system performance. Figure 4 In (a), as V increases, the system queue backlog gradually increases, leading to decreased queue stability. Meanwhile, in Figure 4 (b) shows that the system energy consumption gradually decreases and tends to stabilize. This is because, in system optimization, the parameters... It can be used to achieve a trade-off between system energy consumption and queue stability, when When the size is relatively small, the system tends to optimize the drift function, i.e., optimize queue stability, at the expense of system utility to some extent. However, when... When the value is sufficiently large, the system primarily considers minimizing energy consumption, task execution and offloading are restricted, and queue backlog increases significantly. Clearly, values ​​that are too small or too large... These values ​​are all detrimental to the system.

[0059] Figure 5 The task arrival rate was studied. Impact on system performance. As the maximum task size increases from 0.2MB to 1MB, such as... Figure 5As shown in (a), the backlog increases sharply once the arrival rate exceeds 0.5 Mbit. However, compared to the OPOA and NCOA methods, this algorithm maintains excellent stability by dynamically adjusting the task allocation among local devices, drones, and satellites to suppress Lyapunov drift. Similarly... Figure 5 In (b), as the maximum number of tasks arriving gradually increases, the system energy consumption curve shows an increasing trend, but the rate of increase gradually slows down. This is because when more tasks arrive, the system needs to allocate more resources to users to prevent the backlog queue from overflowing, thus increasing the system's energy consumption. However, due to limited communication and computing resources, as the amount of data arriving continues to increase, the system cannot meet the data processing requests of all users and will gradually reach the system's processing limit, causing the system's energy consumption to gradually stabilize.

[0060] Traditional terrestrial networks suffer from poor coverage, limited resources, and insufficient emergency response capabilities. In contrast, SAGIN offers advantages such as wide-area coverage, rapid and flexible deployment, and freedom from geographical limitations. Therefore, SAGIN has become an indispensable tool for communication assurance, disaster prevention, and emergency mitigation, serving as a crucial supplement to terrestrial networks. Low Earth Orbit (LEO) satellites further expand the limited coverage of air-to-ground networks, effectively meeting the needs for full coverage, all-weather, and all-day communication. Meanwhile, UAVs (Unmanned Aerial Vehicles) provide a rapidly deployable solution for delivering on-demand communication and adaptive computing resources, enabling high-speed, dynamic network capabilities. Furthermore, Backscatter Communication (BackCom) offers significant advantages in terms of low cost and low power consumption. Specifically, a GU equipped with a backscatter circuit can modulate the incident signal load impedance, transmitting its own information onto the reflected signal to the target node with minimal energy consumption, achieving low-power passive communication and low-cost deployment. To overcome these limitations, UAVs can be used as mobile RF power sources, leveraging their high mobility, flexible deployment capabilities, and adaptive altitude control to ensure reliable line-of-sight (LoS) connections with GUs. Their controllable trajectories and hovering capabilities enable dynamic adjustments to network topology and communication range. Therefore, integrating BackCom into SAGIN provides a viable solution to energy-constrained issues in emergency communications.

[0061] This invention aims to address the problems in existing technologies, such as insufficient communication endurance due to limited energy of terrestrial IoT devices in power emergency scenarios where communication facilities are damaged, and the difficulty in balancing energy efficiency and real-time task processing due to a lack of coordination among air-space-ground network resources. By introducing and optimizing airborne-assisted backscatter communication in an integrated air-space-ground network, energy-constrained terrestrial power devices can upload data in a low-power manner, fundamentally solving the problems of excessive energy consumption and short battery life in traditional active communication. This ensures the long-term operational capability of the network in emergency situations and achieves sustainable emergency communication under low power consumption. Employing an online control framework based on Lyapunov optimization, decisions can be made based on real-time network status without needing to predict future task arrivals and channel changes. By adjusting control parameters, an optimal balance can be flexibly achieved between total system energy consumption and task processing queue latency, avoiding the shortcomings of traditional solutions that easily fall into local optima or queue out of control, and achieving a dynamic global optimal trade-off between energy efficiency and latency.

[0062] The proposed alternating optimization algorithm effectively addresses non-convex optimization problems involving air-space-ground resources under multivariate coupling. It coordinates the scheduling of various resources such as time, power, and computation, significantly improving the overall resource utilization of the system. Under the same conditions, it supports the timely processing of a larger number of computational tasks and enhances resource coordination efficiency in complex dynamic environments. Combining the flexible line-of-sight coverage of UAVs with the wide-area reliable backhaul of satellites, a multi-layered computing architecture is constructed. When the UAV is overloaded, tasks can be automatically offloaded to satellites, avoiding single-point overload, improving the system's resilience to sudden tasks and large-scale computations, and enhancing the network's reliability and adaptability.

[0063] Traditional terrestrial networks suffer from poor coverage, limited resources, and insufficient emergency response capabilities. In contrast, SAGIN offers advantages such as wide-area coverage, rapid and flexible deployment, and freedom from geographical limitations. Therefore, SAGIN has become an indispensable tool for communication assurance, disaster prevention, and emergency mitigation, serving as a crucial supplement to terrestrial networks. Low Earth Orbit (LEO) satellites further expand the limited coverage of air-to-ground networks, effectively meeting the needs for full coverage, all-weather, and all-day communication. Meanwhile, UAVs (Unmanned Aerial Vehicles) provide a rapidly deployable solution for delivering on-demand communication and adaptive computing resources, enabling high-speed, dynamic network capabilities. Furthermore, Backscatter Communication (BackCom) offers significant advantages in terms of low cost and low power consumption. Specifically, a GU equipped with a backscatter circuit can modulate the incident signal load impedance, transmitting its own information onto the reflected signal to the target node with minimal energy consumption, achieving low-power passive communication and low-cost deployment. To overcome these limitations, UAVs can be used as mobile RF power sources, leveraging their high mobility, flexible deployment capabilities, and adaptive altitude control to ensure reliable line-of-sight (LoS) connections with GUs. Their controllable trajectories and hovering capabilities enable dynamic adjustments to network topology and communication range. Therefore, integrating BackCom into SAGIN provides a viable solution to energy-constrained issues in emergency communications.

[0064] Based on the same inventive concept, this invention also provides an integrated air-space-ground network resource management system, comprising: The construction module is used to build an integrated air-space-ground network for the target power system. The network includes a ground layer, an air layer, and a space layer. The ground layer consists of multiple ground devices (GUs), the air layer is equipped with unmanned aerial vehicles (UAVs), and the space layer is equipped with a single low-Earth orbit (LEO) satellite to undertake offloading computing tasks assigned by the UAVs.

[0065] The computing module is used for each GU to continuously collect real-time status data of the target power system according to the first task queue. The UAV actively transmits radio frequency carrier signals to the GU at a preset frequency. After receiving the radio frequency carrier signals, the GU converts the radio frequency carrier signals into electrical energy storage, modulates the real-time status data onto the radio frequency carrier signals, and backscatters the modulated real-time status data back to the corresponding UAV to complete the task offloading. The UAV decodes the real-time status data modulated by the corresponding GU according to the second task queue, performs the computing task, and transmits the computing results back to the corresponding GU.

[0066] The allocation module is used by the central controller to collect network status information of GU and UAV, air-to-ground channel status of UAV and GU, and air-to-space channel status of UAV and LEO in real time. Based on all the collected information, with the goal of minimizing the average energy consumption of network operation and with the stability of the two task queues and the stability of GU power as constraints, a stochastic optimization function is established for the time allocation, task offloading allocation and computing resource allocation of the network. The stochastic optimization function is solved using the Lyapunov drift penalty function method to obtain the optimal resource allocation scheme among GU, UAV and LEO.

[0067] This invention also provides a computer device. At the hardware level, the computer device includes a processor, an internal bus, a network interface, memory, and non-volatile memory, and may also include other hardware required for various services. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to implement the aforementioned integrated air-space-ground network resource management method.

[0068] The present invention also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described integrated air-space-ground network resource management method.

[0069] Specific limitations regarding the computing system for the integrated air-space-ground network resource management method can be found in the limitations of the integrated air-space-ground network resource management method described above, and will not be repeated here. Each module in the aforementioned integrated air-space-ground network resource management system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0070] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification. Furthermore, the above embodiments only illustrate several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for integrated air-space-ground network resource management, characterized in that, include: Construct an integrated air-space-ground network for the target power system, the network comprising a ground layer, an air layer, and a space layer; The ground layer consists of multiple ground-based equipment (GUs), the air layer is equipped with unmanned aerial vehicles (UAVs), and the space layer is equipped with a single low-Earth orbit (LEO) satellite to undertake offloading computing tasks assigned by the UAVs. Each GU continuously collects real-time status data of the target power system according to the first task queue; the UAV actively transmits radio frequency carrier signals to the GU at a preset frequency. After receiving the radio frequency carrier signals, the GU converts the radio frequency carrier signals into electrical energy storage, modulates the real-time status data onto the radio frequency carrier signals, and backscatters the modulated real-time status data back to the corresponding UAV to complete the task unloading. The UAV decodes and receives the real-time status data modulated by the corresponding GU in the form of the second task queue, performs the calculation task, and transmits the calculation result back to the corresponding GU. The central controller collects network status information of GU and UAV, air-to-ground channel status of UAV and GU, and air-to-space channel status of UAV and LEO in real time. Based on all the collected information, with the goal of minimizing the average energy consumption of network operation and with the stability of the two task queues and the stability of GU power as constraints, a stochastic optimization function is established for the time allocation, task offloading allocation and computing resource allocation of the network. The stochastic optimization function is solved using the Lyapunov drift penalty function method to obtain the optimal resource allocation scheme among GU, UAV and LEO.

2. The integrated air-space-ground network resource management method according to claim 1, characterized in that, The real-time status data includes system fault information, system equipment monitoring data, and sensing data of the system's environment; the network status information of the GU includes the device's remaining energy, the length of the first task queue, and the number of new tasks arriving; the network status information of the UAV includes the load status of the second task queue, the computing resource occupancy, and its own remaining energy status; the air-to-ground channel status corresponds to the link attributes between the UAV and the GU, including channel gain, signal attenuation, and interference level; the air-space channel status corresponds to the link attributes between the UAV and the LEO, including link bandwidth, propagation delay, and signal stability; the GU's power stability specifically means that the power collected by the GU is greater than the power it consumes.

3. The integrated air-space-ground network resource management method according to claim 1, characterized in that, The constraints also include: network transmission rate constraints, reflection coefficient constraints, computational resource constraints for GU and UAV, and air-to-space link delay constraints. The network transmission rate includes the air-to-ground link transmission rate between UAV and GU and the air-to-space link transmission rate between UAV and LEO, and the transmission rate must meet the data transmission requirements of the corresponding link. The reflection coefficient is the signal reflection parameter when GU modulates real-time state data, and its value range is [0,1]. The computational resource constraints for GU and UAV include: the local computation of GU does not exceed its own computational resource limit, and the local computation of UAV does not exceed the computational resource limit of its mobile edge computing (MEC) server. The air-to-space link delay constraint means that the sum of the transmission delay of UAV offloading tasks to LEO and the task processing delay of LEO does not exceed a preset threshold.

4. The integrated air-space-ground network resource management method according to claim 3, characterized in that, The method of solving the stochastic optimization function using the Lyapunov drift penalty function to obtain the optimal resource allocation scheme among GU, UAV, and LEO specifically includes: using the Lyapunov drift penalty function method, by introducing control parameters... V The long-term stochasticity of the stochastic optimization function is transformed into a deterministic optimization problem that depends only on the current time slot state, and then solved using an alternating optimization algorithm. The non-convex part of the air-to-ground link transmission rate with respect to the reflection coefficient is linearized using the concave-convex process CCCP technique, and the reflection coefficient, task offloading ratio, and computational resource allocation are solved. Based on the solution, the time allocation for energy harvesting, backscatter communication, task processing, and data transmission is further optimized, ultimately obtaining the optimal resource allocation scheme among the GU, UAV, and LEO, which is then distributed to each device in the GU, UAV, and LEO for execution. The control parameters... V Used to balance the average energy consumption of network operation with the stability of task queues.

5. The integrated air-space-ground network resource management method according to claim 1, characterized in that, The GU is equipped with an energy harvesting module and a backscatter circuit. The energy harvesting module converts the received radio frequency carrier signal into electrical energy for storage, and the backscatter circuit modulates real-time status data onto the received radio frequency carrier signal. The UAV is equipped with a carrier generator and a mobile edge computing (MEC) server. The carrier generator transmits radio frequency carrier signals, and the MEC server performs local computing tasks.

6. The integrated air-space-ground network resource management method according to claim 3, characterized in that, The local computation of the GU includes: dynamically adjusting the computation frequency through dynamic voltage and frequency scaling (DVFS) technology, and determining the amount of locally processed data within time slot t based on its own computational resource limit.

7. The integrated air-space-ground network resource management method according to claim 1, characterized in that, The task offloading allocation includes the ratio of local computation of the GU to tasks offloaded to the corresponding UAV, and the ratio of local computation of the UAV to tasks offloaded to the corresponding LEO. The computation tasks adopt a partial offloading method, which is divided into a local execution part, an offloaded execution part to the UAV, and an offloaded execution part to the LEO, and the three are executed in parallel.

8. A space-air-ground integrated network resource management system, characterized in that, include: A construction module is used to build an integrated air-space-ground network for the target power system, the network comprising a ground layer, an air layer, and a space layer; The ground layer consists of multiple ground-based equipment (GUs), the air layer is equipped with unmanned aerial vehicles (UAVs), and the space layer is equipped with a single low-Earth orbit (LEO) satellite to undertake offloading computing tasks assigned by the UAVs. The computing module is used for each GU to continuously collect real-time status data of the target power system according to the first task queue; the UAV actively transmits radio frequency carrier signals to the GU at a preset frequency. After receiving the radio frequency carrier signals, the GU converts the radio frequency carrier signals into electrical energy storage, modulates the real-time status data onto the radio frequency carrier signals, and backscatters the modulated real-time status data back to the corresponding UAV to complete the task unloading. The UAV decodes and receives the real-time status data modulated by the corresponding GU in the form of the second task queue, performs the calculation task, and transmits the calculation result back to the corresponding GU. The allocation module is used by the central controller to collect network status information of GU and UAV, air-to-ground channel status of UAV and GU, and air-to-space channel status of UAV and LEO in real time. Based on all the collected information, with the goal of minimizing the average energy consumption of network operation and with the stability of the two task queues and the stability of GU power as constraints, a stochastic optimization function is established for the time allocation, task offloading allocation and computing resource allocation of the network. The stochastic optimization function is solved using the Lyapunov drift penalty function method to obtain the optimal resource allocation scheme among GU, UAV and LEO.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is loaded by the processor, it is able to perform the steps of the method according to any one of claims 1 to 7.

Images