A space-air-ground communication resource multi-objective optimization method based on a pareto optimal solution set

Abstract

The application relates to the field of communication and is a space-air-ground communication resource multi-target optimization method based on a Pareto optimal solution set, which solves the problem of unreasonable communication resource allocation schemes in the prior art. The application relates to a ground Internet of Things terminal, a UAV and a low-orbit satellite, and comprises the following steps: 1. calculating uplink channel capacity from the ground Internet of Things terminal to the low-orbit satellite; 2. optimizing a target function modeling: taking the minimum uplink channel capacity of each category of Internet of Things terminal in step 1 as an optimization target; 3. iterative optimization based on an evolutionary algorithm framework; the application flexibly designs the deployment position of the UAV and the bandwidth allocation scheme among the Internet of Things devices, adopts a directional resource adjustment scheme, and optimizes the uplink channel capacity of each category of ground Internet of Things terminal through the UAV relay to the low-orbit satellite in a fair manner, so that the uplink channel capacity among the same type of Internet of Things devices is as equal as possible and the uplink channel capacity of each category of Internet of Things device is as large as possible.

Description

Technical Field

[0001] This invention relates to the field of communications, and in particular to a multi-objective optimization method for air-space-ground communication resources based on Pareto optimal solution sets. Background Technology

[0002] Driven by the need for the Internet of Things (IoT), the application and deployment of IoT devices are experiencing an unprecedented boom. For IoT devices used for data collection, stable network coverage and sufficient channel speeds are essential, as the collected data needs to be transmitted to a data processing center. However, current communication systems relying on terrestrial base stations as relays cannot achieve global signal coverage. For example, data transmission is limited in areas such as oceans, deserts, and regions where terrestrial network infrastructure is damaged or insufficient.

[0003] To address the aforementioned issues, integrated air-space-ground networks have emerged as a mainstream solution. This approach merges different types of network resources—terrestrial, aerial, and satellite—into a unified network system to achieve seamless global connectivity and information transmission. This integrated network architecture aims to improve the efficiency, stability, and coverage of communication and information transmission to meet the ever-increasing demands for data transmission. Currently, most integrated air-space-ground network architectures consist of terrestrial IoT terminal devices, unmanned aerial vehicles (UAVs), and low-Earth orbit (LEO) satellites.

[0004] To achieve signal coverage for terrestrial IoT terminals, the key technological challenge lies in system resource allocation to optimize the overall system's channel capacity. However, existing technologies still suffer from several issues:

[0005] 1) It ignores the fairness of communication resources among IoT devices. Existing methods optimize resource allocation only by targeting the sum of the channel capacity of all IoT devices, which may lead to large differences in channel capacity among IoT devices and very small channel capacity for some IoT devices;

[0006] 2) It ignores the differences in channel capacity requirements among different types of IoT devices. Because different types of IoT devices collect different types of data, such as video data, audio data, infrared signal data, and radar signal data, their channel capacity requirements also differ.

[0007] 3) It ignores the time-varying nature of the channel capacity requirements of IoT devices. The channel capacity requirements of IoT devices are not constant and may change with the rate at which the data is collected.

[0008] There is an urgent need for a new method to optimize air-space-ground communication resources that can solve the above problems. Summary of the Invention

[0009] This invention proposes a multi-objective optimization method for air-space-ground communication resources based on Pareto optimal solution sets, which solves the problem of unreasonable communication resource allocation schemes in the prior art.

[0010] The technical solution of this invention is implemented as follows: a multi-objective optimization method for air-space-ground communication resources based on Pareto optimal solution sets, involving terrestrial IoT terminals, airborne relay networks, and satellite networks; the airborne relay network is a high-altitude platform device, which can be a drone or an aircraft; the satellite network can be composed of low-Earth orbit satellites, characterized by the following steps: 1. Calculating the uplink channel capacity from the terrestrial IoT terminal to the low-Earth orbit satellite; 2. Optimizing the objective function model: taking the minimum uplink channel capacity of each type of IoT terminal in step 1 as the objective to be optimized; 3. Iterative optimization based on an evolutionary algorithm framework;

[0011] 3.1 Set algorithm parameters: iteration termination condition, population size, neighborhood size; 3.2 Initialize the external population EP set, i.e., External... Population, and the weight vector in the neighborhood; 3.3, Randomly initialize the first generation population: i.e., the deployment location of the UAV, the bandwidth allocation ratio of different types of ground IoT terminals, and the bandwidth allocation ratio within the same type of ground IoT terminal; 3.4 Initialize the global optimal solution based on the first generation population obtained in 3.3; 3.5 Obtain a new resource allocation scheme: for each individual, randomly Gaussian mutate the overall resource classification ratio to obtain a scheme for adjusting resources within the same type of ground IoT terminal to terminals with low channel capacity; in addition, the adjustment amount in each iteration will dynamically decrease exponentially with the number of iterations; 3.6 Determine the weighted Chebyshev distance between the new resource allocation scheme and the global optimal solution. If the distance is less than that of the global optimal solution individual, then replace the global optimal solution individual with the new resource allocation scheme; 3.7 Add the new resource allocation scheme to the external population EP set and continuously filter out dominated solutions; 3.8 Until the iteration termination condition is met, the iteration ends, and the output contains a set of multiple mutually non-dominated feasible solutions.

[0012] Further, step one specifically refers to: 1.1 Setting communication system parameters, including the transmit power of the ground IoT terminal, the forwarding power of the UAV, and the available bandwidth of the system; 1.2 Obtaining the deployment location and type of the ground IoT terminal; 1.3 Constructing an uplink channel model from the ground IoT terminal to the UAV; 1.4 Constructing an uplink channel model from the UAV to the low-Earth orbit satellite; 1.5 Calculating the channel capacity from each ground IoT terminal to the low-Earth orbit satellite by combining Shannon's formula and the signal relay amplification principle.

[0013] Furthermore, the minimum uplink channel capacity mentioned in step two specifically refers to the minimum uplink channel capacity of each category of terrestrial IoT terminals; by adopting the network layered transmission combined with the amplification and forwarding principle, the calculation function of the theoretical value of the minimum uplink channel capacity of different categories of terrestrial IoT terminals is modeled and calculated.

[0014] Step 3.5 specifically refers to: a) adjusting resource allocation among different types of IoT terminals, including bandwidth allocation ratios and drone deployment locations; the method of implementation is to adjust using random Gaussian mutation; b) adjusting the resource occupancy of different IoT devices within the same type of IoT terminal, including bandwidth redistribution ratios and drone deployment locations; the method of implementation is to transfer resources from devices with high channel capacity to devices with low channel capacity, including bandwidth transfer and directional adjustment of drone locations; c) continuously iterating the above steps, and the adjustment amount of each iteration dynamically and exponentially decreases with the number of iterations.

[0015] Furthermore, it also includes step four: flexible adjustment of resource allocation: 4.1, obtain the channel capacity requirements of different types of ground IoT terminals in the current scenario; 4.2, find feasible solutions that match the channel capacity requirements from the set of feasible solutions obtained in step three; 4.3, obtain the deployment location of the UAV, the bandwidth ratio occupied by each type of ground IoT terminal, and the bandwidth ratio occupied by each terminal in the same type of ground IoT terminal based on the matched feasible solutions.

[0016] Furthermore, the new resource allocation scheme in step 3.5 can also be obtained through genetic algorithms, particle swarm optimization algorithms, and immune algorithms.

[0017] This invention discloses a multi-objective optimization method for air-to-ground communication resources based on Pareto optimal solution sets. By flexibly designing the deployment locations of UAVs and bandwidth allocation schemes among various IoT devices, and employing a directional resource adjustment scheme, it fairly optimizes the uplink channel capacity of various terrestrial IoT terminals relayed to low-Earth orbit satellites via UAVs. This ensures that the uplink channel capacity among similar IoT devices is as equal as possible, and that the uplink channel capacity of each type of IoT device is maximized. Furthermore, this invention can generate resource allocation schemes applicable to various demand scenarios in one go, enabling rapid adjustments to resource allocation to meet the channel capacity requirements of terrestrial IoT terminals in different locations when demands change. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are 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.

[0019] Figure 1 Diagram of an integrated air-space-ground uplink network model;

[0020] Figure 2 Flowchart for generating feasible solutions for resource allocation; Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] This invention discloses a multi-objective optimization method for air-space-ground communication resources based on Pareto optimal solution sets, involving terrestrial IoT terminals, airborne relay networks, and satellite networks; the airborne relay network is a high-altitude platform device, which can be a UAV or an aircraft; the satellite network can be composed of low-Earth orbit satellites. The method is characterized by the following steps: 1. Calculating the uplink channel capacity from the terrestrial IoT terminal to the low-Earth orbit satellite; 2. Modeling the objective function: using the minimum uplink channel capacity of each type of IoT terminal in step 1 as the objective to be optimized; 3. Iterative optimization based on an evolutionary algorithm framework: the evolutionary algorithm can be a multi-objective optimization algorithm, such as NSGA-II (Nondominated Sorting Genetic Algorithm II); SPEA2 (Strength Pareto Evolutionary Algorithm 2), etc.

[0023] 3.1 Set algorithm parameters: iteration termination condition, population size, neighborhood size; 3.2 Initialize the external population EP set, i.e., External... Population, and the weight vector in the neighborhood; 3.3, Randomly initialize the first generation population: i.e., the deployment location of the UAV, the bandwidth allocation ratio of different types of ground IoT terminals, and the bandwidth allocation ratio within the same type of ground IoT terminal; 3.4 Initialize the global optimal solution based on the first generation population obtained in 3.3; 3.5 Obtain a new resource allocation scheme: for each individual, randomly Gaussian mutate the overall resource category ratio to obtain a scheme for adjusting resources within the same type of ground IoT terminal to terminals with low channel capacity; in addition, the adjustment amount in each iteration will dynamically decrease exponentially with the number of iterations. The individual is the name for a solution in the optimization algorithm, i.e., a resource allocation scheme, including the bandwidth allocation ratio and the location deployment of the UAV; 3.6 Determine the weighted Chebyshev distance between the new resource allocation scheme and the global optimal solution. If the distance is less than that of the global optimal solution individual, then replace the global optimal solution individual with the new resource allocation scheme; 3.7 Add the new resource allocation scheme to the external population EP set and continuously filter out dominated solutions; 3.8 Until the iteration termination condition is met, the iteration ends, and the output contains a set of multiple mutually non-dominated feasible solutions.

[0024] Further, step one specifically refers to: 1.1 Setting communication system parameters, including the transmit power of the ground IoT terminal, the forwarding power of the UAV, and the available bandwidth of the system; 1.2 Obtaining the deployment location and type of the ground IoT terminal; 1.3 Constructing an uplink channel model from the ground IoT terminal to the UAV; 1.4 Constructing an uplink channel model from the UAV to the low-Earth orbit satellite; 1.5 Calculating the channel capacity from each ground IoT terminal to the low-Earth orbit satellite by combining Shannon's formula and the signal relay amplification principle.

[0025] Furthermore, the minimum uplink channel capacity mentioned in step two specifically refers to the minimum uplink channel capacity of each category of ground IoT terminals; the calculation function of the theoretical value of the minimum uplink channel capacity of different categories of ground IoT terminals is modeled and calculated by adopting the network layered transmission combined with the amplification and forwarding principle. Specifically: (1) Channel from ground IoT terminal to airborne UAV: ​​Shannon formula is used for direct modeling; (2) Channel from airborne UAV to low-Earth orbit satellite: Shannon formula is used for direct modeling; (3) The channel capacity from ground IoT terminal to low-Earth orbit satellite can be obtained from the two-layer model of 1 and 2.

[0026] Step 3.5 specifically refers to: a) adjusting resource allocation among different types of IoT terminals, including bandwidth allocation ratios and drone deployment locations; this is achieved through random Gaussian mutation to diversify resource occupancy among different categories, thereby increasing the diversity of the final Pareto solution; b) adjusting resource occupancy among different IoT devices within the same type of IoT terminal, including bandwidth redistribution ratios and drone deployment locations; this is achieved by transferring resources from devices with high channel capacity to devices with low channel capacity, including bandwidth transfer and directional adjustment of drone locations, to ensure that the channel capacity of devices within the same type of IoT device is as fair and equal as possible; c) continuously iterating the above steps, with the adjustment amount dynamically and exponentially decreasing with each iteration, to converge to the optimal solution.

[0027] Furthermore, it also includes step four: flexible adjustment of resource allocation: 4.1, obtain the channel capacity requirements of different types of ground IoT terminals in the current scenario; 4.2, find feasible solutions that match the channel capacity requirements from the set of feasible solutions obtained in step three; 4.3, obtain the deployment location of the UAV, the bandwidth ratio occupied by each type of ground IoT terminal, and the bandwidth ratio occupied by each terminal in the same type of ground IoT terminal based on the matched feasible solutions.

[0028] Furthermore, the new resource allocation scheme in step 3.5 can also be obtained through genetic algorithms, particle swarm optimization algorithms, and immune algorithms.

[0029] The entire resource optimization modeling and simulation process is as follows:

[0030] Channel modeling: The channels from the ground-based IoT terminal to the aerial drone and from the drone to the low-Earth orbit satellite are modeled separately. Modeling is performed using a unit-distance gain approach to obtain the signal-to-noise ratio (SNR) and channel capacity for these two channels. Taking a Class A IoT device as an example, the modeling formula is as follows:

[0031]

[0032]

[0033]

[0034]

[0035]

[0036] Among them, g O It is the gain per unit distance. It is the distance from the nth IoT device in category A to the drone. This is the channel gain from the nth Class A IoT device to the drone; d U→L It is the distance from the drone to the low-Earth orbit satellite. It is the channel gain from the UAV to the low-Earth orbit satellite; P I It refers to the signal transmission power of IoT devices. This represents the bandwidth allocation for the nth Class A IoT device, where N0 is the Gaussian white noise power. P is the signal-to-noise ratio of the channel from the nth Class A IoT device to the drone; U N represents the signal transmission power of the drone, and N is the number of IoT devices connected to the drone. It is the signal-to-noise ratio of the channel through which the drone relays the signal from the nth IoT device to the low-Earth orbit satellite; finally, It is the channel capacity from the nth Class A IoT device to the low-Earth orbit satellite.

[0037] Objective function modeling: The objective is to maximize the minimum channel capacity among various IoT devices. The formula is as follows:

[0038]

[0039] L = [L U ,K,k A k B ...]

[0040] Here, the category of IoT devices can be A, B, and others; L represents the resource allocation content to be optimized, including L U This indicates the deployment location of the drone, and K represents the bandwidth allocation ratio for different types of IoT devices. A and k B Equals represent the bandwidth allocation ratio among various devices within a certain type of Internet of Things (IoT) device;

[0041] The mechanism for generating new solutions in evolutionary algorithms:

[0042] This involves adjusting resource allocation among different types of IoT terminals, including bandwidth allocation ratios and drone deployment locations. This is achieved through random Gaussian mutation. The aim is to diversify resource allocation across different categories, thereby increasing the diversity of the final Pareto solution; the formula is as follows:

[0043] x′ u =x u +e x y′ u =y u +e y

[0044]

[0045]

[0046] Where, x′ u and y′ u These are the coordinates after the drone's position has been adjusted, x u and y u These are the coordinates before adjustment; e x and e y It is the adjustment amount of the position, which is randomly determined by a Gaussian distribution; K′ A and K′ B This refers to the bandwidth usage of IoT devices of types A and B. and It is the bandwidth adjustment amount, and both the adjustment amount and the adjustment direction are determined randomly.

[0047] b. Adjusting the resource allocation of different IoT devices within the same type of IoT terminal: This includes bandwidth redistribution and drone deployment location. The method involves transferring resources from devices with high channel capacity to devices with low channel capacity, including bandwidth transfer and directional adjustment of drone locations. The goal is to ensure that the channel capacity of devices within the same type of IoT is as fair and equal as possible; the formula is described below:

[0048]

[0049] d x =(x Imin -x′ u ) / dis x d y =(y lmin -y′ u ) / dis y

[0050]

[0051]

[0052]

[0053]

[0054] Regarding the adjustment of the drone's position: x″ x and y″ u These are the adjusted drone coordinates. and It is the coordinate adjustment vector, including the adjustment direction d. x d y and adjusting scalar and dis x and dis yIt is the horizontal and vertical distance between the drone and the IoT device, x Imin and y Imin The x and y axes represent the minimum channel capacity for IoT devices; for bandwidth allocation adjustments: k Amin′ and k Amax′ This refers to the bandwidth occupancy of IoT devices with the smallest and largest channel capacities among Class A IoT devices, after adjustment. and This is the bandwidth adjustment amount for Class A IoT devices; similarly, k Bmin′ k Bmax′ , This is a partial definition for Class B IoT devices;

[0055] c iterates through the above steps, with the adjustment amount in each iteration dynamically and exponentially decreasing with the number of iterations, aiming to converge to the optimal solution. The step size r of the iterations can be decreased using the following formula:

[0056] r = 0.8 epoch *r init

[0057] Where epoch is the current iteration number, r init It is the initial step size, defined as 1;

[0058] This invention discloses a multi-objective optimization method for air-to-ground communication resources based on Pareto optimal solution sets. By flexibly designing the deployment locations of UAVs and bandwidth allocation schemes among various IoT devices, and employing a directional resource adjustment scheme, it fairly optimizes the uplink channel capacity of various terrestrial IoT terminals relayed to low-Earth orbit satellites via UAVs. This ensures that the uplink channel capacity among similar IoT devices is as equal as possible, and that the uplink channel capacity of each type of IoT device is maximized. Furthermore, this invention can generate resource allocation schemes applicable to various demand scenarios in one go, enabling rapid adjustments to resource allocation to meet the channel capacity requirements of terrestrial IoT terminals in different locations when demands change.

[0059] Of course, those skilled in the art should be able to make various corresponding changes and modifications based on the present invention without departing from its spirit and essence, but all such corresponding changes and modifications should fall within the protection scope of the appended claims.

Claims

1. A method for multi-objective optimization of space-air-ground communication resources based on a Pareto optimal solution set, involving a ground Internet of Things terminal, an aerial relay network, and a satellite network; the aerial relay network can be a high-altitude platform device; the satellite network can be composed of low-orbit satellites, characterized in that: The method comprises the following steps: I. Calculate the uplink channel capacity of the ground Internet of Things terminal to the low-orbit satellite; II. Optimize the objective function modeling: take the minimum uplink channel capacity of each category of Internet of Things terminal in step I as the optimization target; III. Iterative optimization based on the framework of evolutionary algorithm: 3.1, set algorithm parameters: iteration termination condition, population size, neighborhood size; 3.2, initialize the external population EP set and the weight vector in the neighborhood; 3.3, randomly initialize the initial population: the deployment position of the unmanned aerial vehicle, the bandwidth allocation ratio of different categories of ground Internet of Things terminals, and the bandwidth allocation ratio of the same category of ground Internet of Things terminals; 3.4, initialize the global optimal solution according to the initial population obtained in 3.3; 3.5, obtain a new resource allocation scheme: for the resource occupation ratio of different categories of terminals, adjust it using random Gaussian mutation; for the resource occupation ratio of terminals in the same category, adjust it by transferring resources to terminals with low channel capacity, thereby obtaining a new resource allocation scheme; in addition, the adjustment amount of each iteration will be dynamically exponentially decayed with the number of iterations; 3.6, judge the weighted Chebyshev distance between the new resource allocation scheme and the global optimal solution, if the distance is less than the global optimal solution, replace the global optimal solution with the new resource allocation scheme; 3.7, add the new resource allocation scheme to the external population EP set, and constantly filter out the dominated solutions; 3.8, until the iteration termination condition is met, the iteration is ended, and a set of multiple non-dominated feasible solutions is output; IV. Flexible adjustment of resource allocation: 4.1, obtain the channel capacity demand of different categories of ground Internet of Things terminals in the current scenario; 4.2, find a feasible solution that matches the channel capacity demand in the feasible solution set obtained in step III; 4.3, obtain the deployment position of the unmanned aerial vehicle, the bandwidth proportion occupied by each category of ground Internet of Things terminal, and the bandwidth proportion occupied by each terminal in the same category of ground Internet of Things terminal according to the matched feasible solution.

2. The space-air-ground communication resource multi-objective optimization method based on the Pareto optimal solution set according to claim 1, wherein step I specifically refers to: 1.1, set the communication system parameters, including the transmit power of the ground Internet of Things terminal, the forward power of the unmanned aerial vehicle, and the system available bandwidth; 1.2, obtain the deployment position and category of the ground Internet of Things terminal; 1.3, construct the uplink channel model of the ground Internet of Things terminal to the unmanned aerial vehicle; 1.4, construct the uplink channel model of the unmanned aerial vehicle to the low-orbit satellite; 1.5, calculate the channel capacity of each ground Internet of Things terminal to the low-orbit satellite by combining the Shannon formula and the signal relay amplification principle.

3. The space-air-ground communication resource multi-objective optimization method based on the Pareto optimal solution set according to claim 2, wherein the minimum uplink channel capacity in step II specifically refers to the minimum uplink channel capacity of each category of ground Internet of Things terminal; The minimum uplink channel capacity of each category of ground Internet of Things terminal is calculated by modeling the calculation function of the theoretical value of the minimum uplink channel capacity of different categories of ground Internet of Things terminals using network hierarchical transmission combined with the amplification and forwarding principle.

4. The method of claim 3, wherein the method is characterized by: Step 3.5 specifically refers to: a. Resource allocation adjustment between different categories of Internet of Things terminals, including bandwidth allocation ratio and deployment location of unmanned aerial vehicles; the method is to adjust through random Gaussian variation; b. Adjustment of resource occupancy of different Internet of Things devices within the same type of Internet of Things terminal, including bandwidth redistribution ratio and deployment location of unmanned aerial vehicles; the method is to transfer resources from high channel capacity devices to low channel capacity devices, including bandwidth transfer and directional adjustment of unmanned aerial vehicle location; c. Continuously iterate the above steps, and the adjustment amount of each iteration will be dynamically exponentially decayed with the number of iterations.

5. The method of claim 4, wherein: The new resource allocation scheme in step 3.5 can also be obtained through genetic algorithm, particle swarm algorithm, and immune algorithm.

Images