A multi-user cooperative optimization low earth orbit satellite handover method and system

Abstract

This invention relates to the field of satellite communication technology and discloses a multi-user collaborative optimization method and system for low-Earth orbit (LEO) satellite handover, comprising the following steps: establishing a non-cooperative game model with users as players and the set of legal handover paths as the action space; constructing a revenue function that integrates path rewards and resource price costs; initializing satellite time slot resource prices and iteration parameters; using dynamic programming backtracking to solve for the optimal user path; updating the excess resource price based on resource occupancy; iteratively optimizing until capacity constraints are met or termination conditions are triggered; distinguishing between conflicting and non-conflicting users; fixing the paths and occupied resources of non-conflicting users; reconstructing the game model for conflicting users and completing a second iteration; merging the optimal paths to form a global solution. This invention solves the problems of frequent handover and resource conflicts in multi-user resource competition in LEO satellite networks, effectively improving the total system reward, shortening system runtime, reducing computational complexity, and demonstrating good feasibility.

Description

Technical Field

[0001] This invention relates to the field of satellite communication technology, specifically to a multi-user collaborative optimization method and system for low-Earth orbit satellite handover. Background Technology

[0002] In recent years, with the rapid development of wireless communication technology, low-Earth orbit (LEO) satellite constellation technology, as a core supporting technology for integrated space-ground communication, has made significant contributions to expanding communication coverage and improving service reliability in complex scenarios. LEO satellite constellations, with their advantages of low latency, wide coverage, and flexible networking, can meet real-time communication needs, compensate for shortcomings in terrestrial network coverage, and provide stable support for scenarios such as connectivity in remote areas and emergency communication. However, while LEO satellite constellation technology has significant advantages, it also faces the challenge of complex handover decisions.

[0003] In low-Earth orbit (LEO) satellite networks, the high-speed motion of LEO satellites causes continuous changes in the link status between users and satellites, resulting in frequent handover processes. Satellite nodes have rigid limitations in their carrying capacity, and in multi-user access scenarios, handover decisions are mutually influential, easily leading to resource contention and conflicts. A comprehensive consideration of multiple attributes and a balance of temporal correlations in handover decisions play a crucial role in ensuring the continuity and stability of communication services, effectively suppressing phenomena such as ping-pong handovers and handover failures. Furthermore, effective management of resource conflicts can improve satellite resource utilization and prevent overall network performance degradation caused by resource contention. Summary of the Invention

[0004] To address the technical problem that existing low-Earth orbit (LEO) satellite handover schemes in multi-user resource contention scenarios struggle to balance satellite user capacity constraints, the temporal correlation of handover sequences, and the comprehensive value of service time and service capacity, resulting in poor service continuity and stability of the communication system, this invention proposes a multi-user collaborative optimization method and system for LEO satellite handover. The technical solution is as follows:

[0005] A multi-user collaborative optimization method for low-Earth orbit satellite handover includes the following steps:

[0006] Step 1: Establish a non-cooperative game model, with each user as a player, define the player's action space as a set of legal switching paths, construct a revenue function that integrates path rewards and resource price costs, and form a multi-user switching decision optimization framework.

[0007] Step 2: Initialize satellite time slot resource prices, set iteration parameters, and use dynamic programming backtracking method to solve the optimal path for each user under the current price to form action combinations; count resource occupancy, identify iteration abnormal scenarios, and if no abnormality occurs, update the excess resource price according to the resource congestion level, and iteratively execute path selection and resource price adjustment until the satellite capacity constraint is met or the conflict resolution condition is triggered.

[0008] Step 3: If the conflict resolution condition is triggered, distinguish between conflicting and non-conflicting users, fix the path of non-conflicting users and lock their occupied resources, and repeat step 2 for conflicting users.

[0009] Step 4: Calculate and output the total reward for all users.

[0010] Furthermore, the communication system of the low-orbit satellite specifically comprises: a system with... Individual users The communication system of the low-Earth orbit satellites loads satellite orbital parameter information onto each satellite in the constellation. , And randomly obtain its latitude range Longitude range within User geographic coordinates , The user set is Set the basic time unit Set the start time slot based on the current access request time. Set the time slot sequence according to the duration of the requested connection. , , ,in, Total number of service periods For the end-of-slot node, the service period covers common Each service period has a unique service period number. ;

[0011] Based on satellite orbital parameter information and user geographic coordinates Obtain the time slot sequence satellite For users pitch angle and distance ,in For each user Computing satellites of The time slot is used as the effective coverage time slot, among which Determine the first effective coverage slot based on the minimum elevation angle threshold. and the last valid coverage slot constituting a satellite For users Available service hours and effective coverage start and end time slot sequence numbers If satellite In the time slot Service users Then it is considered a satellite Service hours are ;

[0012] Computing satellites In the time slot For users Continuous service duration:

[0013] ,

[0014] in, Calculated as the minimum service duration threshold , , To minimize the number of consecutive service periods during switching; for users , obtain arrive The process involves a set of all satellites with valid coverage time slots. , can be represented as:

[0015] ,

[0016] in, and will The satellites in the sequence are ordered according to the end slot number. The set of satellites effectively covering time slots after ascending sorting can be represented as follows:

[0017] ,

[0018] in, The total number of elements in the set; iterating through the set. ,Pick The Middle satellite ,satellite Switching candidate set , can be represented as:

[0019] ,

[0020] in, To switch satellites The initial service slot sequence number, For users In time slot number Can switch to satellite The symbol is expressed as:

[0021] ,

[0022] Can switch to satellite The condition is the sequence number of the end slot. for or continuous service duration Meet the minimum service duration threshold ;

[0023] Calculate free space propagation loss:

[0024] ,

[0025] in, For the system carrier frequency, Given the speed of light; calculate the received signal power:

[0026] ,

[0027] in, Satellite launch power, Given the total antenna gain; calculate the signal-to-noise ratio in the logarithmic domain:

[0028] ,

[0029] in, For noise power; convert the logarithmic domain signal-to-noise ratio to the linear domain signal-to-noise ratio:

[0030] ,

[0031] Calculate the spectral efficiency using Shannon's formula:

[0032] ,

[0033] Calculate the transmission rate:

[0034] ,

[0035] in, For satellite bandwidth, calculate all users exist Average rate of visible service satellites in the time slot:

[0036] ,

[0037] in, For users exist The set of available satellites for a time slot, For users exist Number of available satellites in a time slot; calculation of users exist Time slot selection satellite Initial service reward for providing the service:

[0038] ,

[0039] Calculate users Service satellite by Switch to Single-step switching reward:

[0040] ,

[0041] in, For users China Satellite exist The transmission rate of a time slot, For users China Satellite exist The remaining service time of the time slot, For satellite The switching time slot; user Switching path Its expression is:

[0042] ,

[0043] in, For path Number of switches, path shared by It consists of satellite-time slots; each satellite-time slot Equivalent to ,in, , express The Middle For users Satellite service Service users The time period is ;

[0044] According to the path Calculate path reward:

[0045] ,

[0046] in, For path The first one is the user satellite services Initial service bonus, For path From the middle service satellites Switch to the service satellites The single-step switching reward; a single satellite can serve a maximum of one user per time slot, and each user is continuously served by a certain satellite in each time slot. Users only switch satellites when the coverage of the currently serving satellite ends.

[0047] Furthermore, in step 1, a non-cooperative game theory model for multi-user handover decision-making is established under the communication system, and a multi-user collaborative handover optimization framework is constructed. This game theory model is defined as follows:

[0048] ,

[0049] With all users in the system For players, every user With the goal of maximizing its own profits, and with the user as the focus The set of all legal switching paths constitutes the action space. The revenue function is a function of the fusion path reward and resource price cost, as shown in the following formula:

[0050] ,

[0051] in, For the first Satellite time slots in the next iteration Resource prices; To exclude players The action combinations of the other players can be represented as:

[0052] ,

[0053] For players When other players' action combinations Given, make the payoff function Maximize Action The best response is:

[0054] ,

[0055] The combination of all players' actions is denoted as , can be represented as:

[0056] ,

[0057] in, For players Selected legal switching path; when action combination Satisfying all players , All are action combinations against other players. The best response, i.e. At that time, all players cannot change their actions. Increase profits This action combination The Nash equilibrium state is reached.

[0058] Furthermore, in step 2, given the resource price, the user solves for the optimal response path using dynamic programming backtracking. The resulting strategy combination reaches Nash equilibrium at this price. The resource occupancy is statistically analyzed, and the price of excess resources is updated. The price update changes the user's revenue function, prompting the user to re-solve for the optimal response based on the new price and enter the next iteration.

[0059] Furthermore, step 2 specifically includes:

[0060] Step 2.1: Set the price adjustment coefficient Satellite time slot resource price matrix and maximum number of iterations ,definition Initialize the iteration counter to the number of iterations. Based on the initial resource price matrix Begin the iterative optimization process; set all satellite time slots. The initial price of the resource is:

[0061] ;

[0062] Step 2.2, for the user Effective satellite set Build a status record table Each satellite Each record contains three core pieces of information: arrival at satellite. Maximum cumulative net income The initial value is set to ; Arrive at satellite Optimal precursor satellite The initial value is set to null; current satellite identifier. ; Calculate users exist Time slot selection satellite Initial service revenue from providing the service:

[0063] ,

[0064] in, For the first Satellite resources during the next iteration Price, satellite update Maximum cumulative net income traverse users All effective coverage satellites Determine the satellite The set of switchable precursor satellites is expressed as:

[0065] ,

[0066] Traverse all switchable precursor satellites Calculate from Switch to Cumulative net income:

[0067] ,

[0068] Update satellite Maximum cumulative net profit and corresponding optimal precursor satellite:

[0069] ,

[0070] in, To reach the satellite Maximum cumulative net income, For satellite The optimal precursor satellite is used to determine the endpoint time slot. Serviceable users The set of satellites, expressed as:

[0071] ,

[0072] exist Select the largest cumulative net income Largest satellite As the best response terminal satellites, users The maximum path reward is ,Right now:

[0073] ,

[0074] by To trace back to the starting point, for the current node ,calculate Start time slot number:

[0075] ,

[0076] in, To track the number of steps, To form The number of satellite-time slots for The satellite-time slot sequence number, , , ; Calculate the end-of-service slot sequence number:

[0077] ,

[0078] Record path The Middle Satellite-time slot: ,renew Increasing until Empty, forming a path Its expression is:

[0079] ;

[0080] Step 2.3, Best Reactions from All Players Form the current iterative action combination Its expression is:

[0081] ,

[0082] Calculation path Rewards: ,

[0083] Statistics Temporary total reward for the next iteration: ,

[0084] Define the matrix of users occupying satellite time slot resources iteration At this time, satellite time slot resources The number of users occupying the space is:

[0085] ,

[0086] in, For indicator functions, when hour ,otherwise Mark the set of resources that exceed the limit:

[0087] ,

[0088] in, For satellite service user capacity, when If the current price is empty, there are no excess resources for price updates, and all satellite time slot resources meet capacity constraints, the best user response path at the current price stops iterating. The path in the file is the user's final path, and the first... Temporary total reward for the next iteration Set as the total system reward : ;

[0089] when Not empty, records each resource exceeding the limit. User set:

[0090] ,

[0091] Calculate the resource impact coefficient:

[0092] ,

[0093] in, for Price for users The maximum path benefit; To temporarily prohibit the use of resources Afterwards, the user The suboptimal path benefit is obtained by using dynamic programming and backtracking.

[0094] Furthermore, if step 2 satisfies any of the following scenarios, the current price adjustment mechanism is determined to be invalid, the iteration is stopped, and step 3 is performed:

[0095] Scenario 1: Price mechanism fails: Excessive resource set exists. Furthermore, the influence coefficients of all resources in this set are zero, i.e. Price updates cannot resolve resource conflicts;

[0096] Scenario 2: Iteration Enters a Loop: Continuous In this iteration, the set of excess resources, the set of users using excess resources, and the temporary total reward remain consistent. The algorithm oscillates and cannot converge autonomously;

[0097] Scenario 3: Reaching the computing power limit: The number of iterations reaches its maximum value. And there are still resources exceeding the limit;

[0098] If the above scenario is not met, update the price of excess resources:

[0099] ,

[0100] Prices for resources not exceeding limits remain unchanged:

[0101] ,

[0102] Update iteration count Return to step 2.3 to execute the next iteration.

[0103] Furthermore, in step 3, users who occupy excessive resources are identified as conflicting users, and the rest are non-conflicting users. The paths of non-conflicting users are fixed and their occupied resources are included in the disabled set. The non-cooperative game model of conflicting users is reconstructed. Price adjustment iteration is reproduced under the constraint of disabled resources. When the subgame converges to no resource excess, the optimal paths of the two types of users are merged to form a globally feasible solution.

[0104] Furthermore, step 3 specifically includes:

[0105] Step 3.1: Divide users into conflicting users and non-conflicting users. The set of conflicting users consists of users who occupy at least one resource exceeding the limit, i.e.:

[0106] ,

[0107] in, Over-limit resources when stopping iteration , To use excessive resources The user set, The union operation is performed on sets; the set of non-conflicting users is defined as the complement of the set of conflicting users.

[0108] ,

[0109] The path of a non-conflicting user at the point of stopping iteration is identified as the final path. This set of final paths for non-conflicting users is denoted as: And extract all satellite time slot resources occupied by these paths to form a set of disabled resources:

[0110] ;

[0111] Step 3.2: Reconstruct the non-cooperative game model of conflicting users:

[0112] ,

[0113] Gather the players as Remove all paths containing disabled resources to create new action space for conflicting users. The corresponding change to the profit function is as follows: Set the maximum number of iterations. Initialize the iteration counter Reconstructing the satellite time slot resource price matrix And set the initial price for conflicting user iterations:

[0114] ,

[0115] in, Under this setting, for Execute the iterative process in step 2.3. When the iteration converges, that is... At that time, the set of optimal paths for conflicting users is:

[0116] ,

[0117] The fixed paths of non-conflicting users are merged with the convergence paths of conflicting users to form the final set of switching paths for all users:

[0118] ,

[0119] Through users Selected path Calculate the reward for the corresponding path And calculate the total reward for all user paths. :

[0120] .

[0121] A multi-user collaborative optimization low-Earth orbit satellite handover system, used to implement any of the methods described above, includes the following modules:

[0122] Game model construction module: used to build non-cooperative game models, define the payoff functions of game players, action space and fusion path rewards and resource price costs, and form a multi-user switching decision optimization framework;

[0123] Price Adjustment Iterative Optimization Module: Used to initialize satellite time slot resource prices and iterative parameters, solve the optimal user path using dynamic programming backtracking method, count resource occupancy and update over-limit resource prices, and perform iterative optimization;

[0124] The conflict resolution and secondary optimization module is used to distinguish between conflicting and non-conflicting users, fix the paths and resources occupied by non-conflicting users, reconstruct the game model of conflicting users and complete the secondary iteration, and output the globally optimal switching solution.

[0125] Beneficial effects

[0126] This invention constructs a game model that integrates service time and capacity, uses a price adjustment iterative optimization mechanism to solve for the optimal response path, and introduces conflict resolution and secondary game strategies. Under the premise of meeting the single satellite single time slot capacity constraint and satellite continuous service requirements, it effectively improves the total system reward, significantly reduces computational complexity, ensures real-time response capability, and adapts to the high-speed dynamic topology change characteristics of low-Earth orbit satellite networks. Attached Figure Description

[0127] Figure 1 Flowchart of a low-Earth orbit satellite handover method optimized for multi-user collaboration;

[0128] Figure 2 A diagram of a low-Earth orbit satellite handover system optimized for multi-user collaboration;

[0129] Figure 3 This is a simulation diagram showing how the total reward varies with the number of users for the three allocation methods in this embodiment of the invention;

[0130] Figure 4 This is a simulation experiment diagram showing the running time of the three allocation methods as a function of the number of users in this embodiment of the invention. Detailed Implementation

[0131] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0132] In this embodiment, a single satellite serves a maximum of one user per time slot. Each user must be continuously served by a specific satellite in each time slot, and users can only switch satellites when the coverage of the currently serving satellite ends. Latitude range is defined. Longitude range Number of users Take respectively Satellite orbital parameter information is Parameter information of Starlink low-Earth orbit satellites, starting time slot Basic unit of time Total number of service hours Minimum elevation angle threshold Switch to the minimum number of consecutive service periods System carrier frequency speed of light Satellite launch power Total antenna gain noise power Satellite bandwidth Price adjustment coefficient Its relationship with the number of users The values ​​correspond one-to-one, when for hour, They are respectively Satellite service user capacity Maximum number of iterations for all users in the game Maximum number of iterations in conflict user game Its application scenario is low-orbit satellite communication systems.

[0133] like Figure 1 As shown, the multi-user collaborative optimization method for low-Earth orbit satellite handover of the present invention includes the following steps:

[0134] Step 1: Establish a non-cooperative game theory model for multi-user handover decision-making under the above communication system, and construct a multi-user collaborative handover optimization framework. This game theory model is defined as follows:

[0135] ,

[0136] With all users in the system For players, every user With the goal of maximizing its own profits, and with the user as the focus The set of all legal switching paths constitutes the action space. The revenue function is a function of the fusion path reward and resource price cost, as shown in the following formula:

[0137] ,

[0138] in, For the first Satellite time slots in the next iteration Resource prices; To exclude players The action combinations of the other players can be represented as:

[0139] ,

[0140] For players When other players' action combinations Given, make the payoff function Maximize Action The best response is:

[0141] ,

[0142] The combination of all players' actions is denoted as , can be represented as:

[0143] ,

[0144] in, For players Selected legal switching path; when action combination Satisfying all players , All are action combinations against other players. The best response, i.e. At that time, all players cannot change their actions. Increase profits This action combination The Nash equilibrium state is reached.

[0145] Step 2: Given a resource price, the user solves the optimal response path using dynamic programming and backtracking. The resulting strategy combination reaches Nash equilibrium at this price. Resource occupancy is statistically analyzed, and the price of excess resources is updated. The price update changes the user's payoff function, prompting the user to re-solve the optimal response based on the new price and enter the next iteration. This process guides the system to gradually avoid resource conflicts and achieve collaborative optimization through price adjustment in multiple iterations.

[0146] Step 2.1: Set the price adjustment coefficient Satellite time slot resource price matrix and maximum number of iterations ,definition Initialize the iteration counter to the number of iterations. Based on the initial resource price matrix Begin the iterative optimization process; set all satellite time slots. The initial price of the resource is:

[0147] ;

[0148] Step 2.2, for the user Effective satellite set Build a status record table Each satellite Each record contains three core pieces of information: arrival at satellite. Maximum cumulative net income The initial value is set to ; Arrive at satellite Optimal precursor satellite The initial value is set to null; current satellite identifier. ; Calculate users exist Time slot selection satellite Initial service revenue from providing the service:

[0149] ,

[0150] in, For the first Satellite resources during the next iteration Price, satellite update Maximum cumulative net income traverse users All effective coverage satellites Determine the satellite The set of switchable precursor satellites is expressed as:

[0151] ,

[0152] Traverse all switchable precursor satellites Calculate from Switch to Cumulative net income:

[0153] ,

[0154] Update satellite Maximum cumulative net profit and corresponding optimal precursor satellite:

[0155] ,

[0156] in, To reach the satellite Maximum cumulative net income, For satellite The optimal precursor satellite is used to determine the endpoint time slot. Serviceable users The set of satellites, expressed as:

[0157] ,

[0158] exist Select the largest cumulative net income Largest satellite As the best response terminal satellites, users The maximum path reward is ,Right now:

[0159] ,

[0160] by To trace back to the starting point, for the current node ,calculate Start time slot number:

[0161] ,

[0162] in, To track the number of steps, To form The number of satellite-time slots for The satellite-time slot sequence number, , , ; Calculate the end-of-service slot sequence number:

[0163] ,

[0164] Record path The Middle Satellite-time slot: ,renew Increasing until Empty, forming a path Its expression is:

[0165] ;

[0166] Step 2.3, Best Reactions from All Players Form the current iterative action combination Its expression is:

[0167] ,

[0168] Calculation path Rewards: ,

[0169] Statistics Temporary total reward for the next iteration: ,

[0170] Define the matrix of users occupying satellite time slot resources iteration At this time, satellite time slot resources The number of users occupying the space is:

[0171] ,

[0172] in, For indicator functions, when hour ,otherwise Mark the set of resources that exceed the limit:

[0173] ,

[0174] in, For satellite service user capacity, when If the current price is empty, there are no excess resources for price updates, and all satellite time slot resources meet capacity constraints, the best user response path at the current price stops iterating. The path in the file is the user's final path, and the first... Temporary total reward for the next iteration Set as the total system reward : ;

[0175] when Not empty, records each resource exceeding the limit. User set:

[0176] ,

[0177] Calculate the resource impact coefficient:

[0178] ,

[0179] in, for Price for users The maximum path benefit; To temporarily prohibit the use of resources Afterwards, the user The suboptimal path benefit is obtained by using dynamic programming and backtracking.

[0180] If any of the following scenarios are met, the current price adjustment mechanism is deemed invalid, the iteration is stopped, and step 3 is executed:

[0181] Scenario 1: Price mechanism fails: Excessive resource set exists. Furthermore, the influence coefficients of all resources in this set are zero, i.e. Price updates cannot resolve resource conflicts;

[0182] Scenario 2: Iteration Enters a Loop: Continuous In this iteration, the set of excess resources, the set of users using excess resources, and the temporary total reward remain consistent. The algorithm oscillates and cannot converge autonomously;

[0183] Scenario 3: Reaching the computing power limit: The number of iterations reaches its maximum value. And there are still resources exceeding the limit;

[0184] If the above scenario is not met, update the price of excess resources:

[0185] ,

[0186] Prices for resources not exceeding limits remain unchanged:

[0187] ,

[0188] Update iteration count Return to step 2.3 to execute the next iteration.

[0189] Step 3: Identify users who occupy excessive resources as conflicting users, and the rest as non-conflicting users. Fix the paths of non-conflicting users and add their occupied resources to the disabled set. Reconstruct the non-cooperative game model of conflicting users. Reproduce the price adjustment iteration under the constraint of disabled resources. When the subgame converges to no resource excess, merge the optimal paths of the two types of users to form a globally feasible solution.

[0190] Step 3.1: Divide users into conflicting users and non-conflicting users. The set of conflicting users consists of users who occupy at least one resource exceeding the limit, i.e.:

[0191] ,

[0192] in, Over-limit resources when stopping iteration , To use excessive resources The user set, The union operation is performed on sets; the set of non-conflicting users is defined as the complement of the set of conflicting users.

[0193] ,

[0194] The path of a non-conflicting user at the point of stopping iteration is identified as the final path. This set of final paths for non-conflicting users is denoted as: And extract all satellite time slot resources occupied by these paths to form a set of disabled resources:

[0195] ;

[0196] Step 3.2: Reconstruct the non-cooperative game model of conflicting users:

[0197] ,

[0198] Gather the players as Remove all paths containing disabled resources to create new action space for conflicting users. The corresponding change to the profit function is as follows: Set the maximum number of iterations. Initialize the iteration counter Reconstructing the satellite time slot resource price matrix And set the initial price for conflicting user iterations:

[0199] ,

[0200] in, Under this setting, for Execute the iterative process in step 2.3. When the iteration converges, that is... At that time, the set of optimal paths for conflicting users is:

[0201] ,

[0202] The fixed paths of non-conflicting users are merged with the convergence paths of conflicting users to form the final set of switching paths for all users:

[0203] ,

[0204] Through users Selected path Calculate the reward for the corresponding path And calculate the total reward for all user paths. :

[0205] .

[0206] The above method fully presents the core process of game model construction, main iterative optimization, and conflict resolution secondary optimization: starting from the beginning, a non-cooperative game model is established, and the satellite time slot resource price is initialized. And set an iteration counter Based on the current price, a dynamic programming backtracking method is used to solve the switching path with the maximum benefit for all users, and resource usage is statistically analyzed and the set of resources exceeding the limit is identified. Determine if the iteration has terminated; if not, update the price. And execute it in a loop; if there are still excessive resources after the main iteration terminates, distinguish conflicting users. With non-conflicting users Fix the paths of non-conflicting users and lock their occupied resources as disabled resources. Reconstruct the game model only for conflicting users and repeat the iterative process. Finally, merge the paths of non-conflicting and conflicting users and output the set of switching paths for all users. And calculate the total system reward. The process is now complete.

[0207] like Figure 2 As shown, the multi-user collaborative optimization low-Earth orbit satellite handover system of the present invention includes the following modules:

[0208] Game model construction module: used to build non-cooperative game models, define the payoff functions of game players, action space and fusion path rewards and resource price costs, and form a multi-user switching decision optimization framework;

[0209] Price Adjustment Iterative Optimization Module: Used to initialize satellite time slot resource prices and iterative parameters, solve the optimal user path using dynamic programming backtracking method, count resource occupancy and update over-limit resource prices, and perform iterative optimization;

[0210] The conflict resolution and secondary optimization module is used to distinguish between conflicting and non-conflicting users, fix the paths and resources occupied by non-conflicting users, reconstruct the game model of conflicting users and complete the secondary iteration, and output the globally optimal switching solution.

[0211] The performance of the method in this invention is compared with that of the prior art's sequential assignment and LP linear programming, such as... Figure 3As shown, with the increase in the number of users, the total reward of the game theory approach of this invention is much higher than that of the sequential allocation approach, and comparable to that of the LP linear programming approach. This advantage is particularly significant when the number of users is large, indicating that this invention can efficiently coordinate multi-user switching decisions and improve system returns. Figure 4 As shown, the running time of sequential allocation is always below the second level. The running time of the game theory method of this invention increases slightly with the number of users but still remains within the second range. In contrast, the running time of LP linear programming increases sharply with the number of users. This indicates that this invention has low computational complexity while ensuring profitability, and is suitable for the real-time requirements of low-orbit satellite networks.

[0212] In summary, this invention proposes a multi-user collaborative optimization method and system for low-Earth orbit (LEO) satellite handover. Combining non-cooperative game theory, dynamic programming backtracking, and conflict resolution techniques, it addresses the highly dynamic topology and multi-user resource competition issues of LEO satellite networks, achieving a comprehensive consideration of multiple attributes and a balance of temporal correlation in handover decisions. Through price adjustment iteration and conflict resolution strategies, this invention reduces computational complexity while satisfying satellite capacity constraints, thereby increasing the overall system reward and ensuring real-time response capabilities. Adapting to the high-speed dynamic characteristics of LEO satellite networks, it possesses significant technical advantages and practical value.

[0213] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A multi-user collaborative optimization method for low-Earth orbit satellite handover, characterized in that, Includes the following steps: Step 1: Establish a non-cooperative game model, with each user as a player, define the player's action space as a set of legal switching paths, construct a revenue function that integrates path rewards and resource price costs, and form a multi-user switching decision optimization framework. Step 2: Initialize satellite time slot resource prices, set iteration parameters, and use dynamic programming backtracking method to solve the optimal path for each user under the current price to form action combinations; count resource occupancy, identify iteration abnormal scenarios, and if no abnormality occurs, update the excess resource price according to the resource congestion level, and iteratively execute path selection and resource price adjustment until the satellite capacity constraint is met or the conflict resolution condition is triggered. Step 3: If the conflict resolution condition is triggered, distinguish between conflicting and non-conflicting users, fix the path of non-conflicting users and lock their occupied resources, and repeat step 2 for conflicting users. Step 4: Calculate and output the total reward for all users.

2. The low-Earth orbit satellite handover method with multi-user collaborative optimization as described in claim 1, characterized in that, The communication system of the low-orbit satellite specifically includes: a system with... Individual users The communication system of the low-Earth orbit satellites loads satellite orbital parameter information onto each satellite in the constellation. , And randomly obtain its latitude range Longitude range within User geographic coordinates , The user set is Set the basic time unit Set the start time slot based on the current access request time. Set the time slot sequence according to the duration of the requested connection. , , ,in, Total number of service periods For the end-of-slot node, the service period covers common Each service period has a unique service period number. ; Based on satellite orbital parameter information and user geographic coordinates Obtain the time slot sequence satellite For users pitch angle and distance ,in For each user Computing satellites of The time slot is used as the effective coverage time slot, among which Determine the first effective coverage slot based on the minimum elevation angle threshold. and the last valid coverage slot constituting a satellite For users Available service hours and effective coverage start and end time slot sequence numbers If satellite In the time slot Service users Then it is considered a satellite Service hours are ; Computing satellites In the time slot For users Continuous service duration: ， in, Calculated as the minimum service duration threshold , , To minimize the number of consecutive service periods during switching; for users , obtain arrive The process involves a set of all satellites with valid coverage time slots. , can be represented as: ， in, and will The satellites in the sequence are ordered according to the end slot number. The set of satellites effectively covering time slots after ascending sorting can be represented as follows: ， in, The total number of elements in the set; iterating through the set. ,Pick The Middle satellite ,satellite Switching candidate set , can be represented as: ， in, To switch satellites The initial service slot sequence number, For users In time slot number Can switch to satellite The symbol is expressed as: ， Can switch to satellite The condition is the sequence number of the end slot. for or continuous service duration Meet the minimum service duration threshold ; Calculate free space propagation loss: ， in, For the system carrier frequency, Given the speed of light; calculate the received signal power: ， in, Satellite launch power, Given the total antenna gain; calculate the signal-to-noise ratio in the logarithmic domain: ， in, For noise power; convert the logarithmic domain signal-to-noise ratio to the linear domain signal-to-noise ratio: ， Calculate the spectral efficiency using Shannon's formula: ， Calculate the transmission rate: ， in, For satellite bandwidth, calculate all users exist Average rate of visible service satellites in the time slot: ， in, For users exist The set of available satellites for a time slot, For users exist Number of available satellites in a time slot; calculation of users exist Time slot selection satellite Initial service reward for providing the service: ， Calculate users Service satellites are Switch to Single-step switching reward: ， in, For users China Satellite exist The transmission rate of a time slot, For users China Satellite exist The remaining service time of the time slot, For satellite The switching time slot; user Switching path Its expression is: ， in, For path Number of switches, path shared by It consists of satellite-time slots; each satellite-time slot Equivalent to ,in, , express The Middle For users Satellite service Service users The time period is ; According to the path Calculate path reward: ， in, For path The first one is the user satellite services Initial service bonus, For path From the middle service satellites Switch to the service satellites The single-step switching reward; a single satellite can serve a maximum of one user per time slot, and each user is continuously served by a certain satellite in each time slot. Users only switch satellites when the coverage of the currently serving satellite ends.

3. The low-Earth orbit satellite handover method with multi-user collaborative optimization as described in claim 2, characterized in that, Step 1 establishes a non-cooperative game theory model for multi-user handover decision-making under the communication system, and constructs a multi-user collaborative handover optimization framework. This game theory model is defined as follows: ， With all users in the system For players, every user With the goal of maximizing its own profits, and with the user as the focus The set of all legal switching paths constitutes the action space. The revenue function is a function of the fusion path reward and resource price cost, as shown in the following formula: ， in, For the first Satellite time slots in the next iteration Resource prices; To exclude players The action combinations of the other players can be represented as: ， For players When other players' action combinations Given, make the payoff function Maximize Action The best response is: ， The combination of all players' actions is denoted as , can be represented as: ， in, For players Selected legal switching path; when action combination Satisfying all players , All are action combinations against other players. The best response, i.e. At that time, all players cannot change their actions. Increase profits This action combination The Nash equilibrium state is reached.

4. The low-Earth orbit satellite handover method with multi-user collaborative optimization as described in claim 1, characterized in that, In step 2, given the resource price, the user solves the optimal response path using dynamic programming backtracking. The resulting strategy combination reaches Nash equilibrium at this price. The resource occupancy is statistically analyzed, and the price of excess resources is updated. The price update changes the user's revenue function, prompting the user to re-solve the optimal response based on the new price and enter the next iteration.

5. The low-Earth orbit satellite handover method with multi-user collaborative optimization as described in claim 3, characterized in that, Step 2 specifically involves: Step 2.1: Set the price adjustment coefficient Satellite time slot resource price matrix and maximum number of iterations ,definition Initialize the iteration counter to the number of iterations. Based on the initial resource price matrix Begin the iterative optimization process; set all satellite time slots. The initial price of the resource is: ； Step 2.2, for the user Effective satellite set Build a status record table Each satellite Each record contains three core pieces of information: arrival at satellite. Maximum cumulative net income The initial value is set to ; Arrive at satellite Optimal precursor satellite The initial value is set to null; current satellite identifier. ; Calculate users exist Time slot selection satellite Initial service revenue from providing the service: ， in, For the first Satellite resources during the next iteration Price, satellite update Maximum cumulative net income traverse users All effective coverage satellites Determine the satellite The set of switchable precursor satellites is expressed as: ， Traverse all switchable precursor satellites Calculate from Switch to Cumulative net income: ， Update satellite Maximum cumulative net profit and corresponding optimal precursor satellite: ， in, To reach the satellite Maximum cumulative net income, For satellite The optimal precursor satellite is used to determine the endpoint time slot. Serviceable users The set of satellites, expressed as: ， exist Select the largest cumulative net income Largest satellite As the best response terminal satellites, users The maximum path reward is ,Right now: ， by To trace back to the starting point, for the current node ,calculate Start time slot number: ， in, To track the number of steps, To form The number of satellite-time slots for The satellite-time slot sequence number, , , ; Calculate the end-of-service slot sequence number: ， Record path The Middle Satellite-time slot: ,renew Increasing until Empty, forming a path Its expression is: ； Step 2.3, Best Reactions from All Players Form the current iterative action combination Its expression is: ， Calculation path Rewards: , Statistics Temporary total reward for the next iteration: , Define the matrix of users occupying satellite time slot resources iteration At this time, satellite time slot resources The number of users occupying the space is: ， in, For indicator functions, when hour ,otherwise Mark the set of resources that exceed the limit: ， in, For satellite service user capacity, when If the current price is empty, there are no excess resources for price updates, and all satellite time slot resources meet capacity constraints, the best user response path at the current price stops iterating. The path in the file is the user's final path, and the first... Temporary total reward for the next iteration Set as the total system reward : ; when Not empty, records each resource exceeding the limit. User set: ， Calculate the resource impact coefficient: ， in, for Price for users The maximum path benefit; To temporarily prohibit the use of resources Afterwards, the user The suboptimal path benefit is obtained by using dynamic programming and backtracking.

6. The low-Earth orbit satellite handover method with multi-user collaborative optimization as described in claim 5, characterized in that, If step 2 satisfies any of the following scenarios, the current price adjustment mechanism is determined to be invalid, the iteration is stopped, and step 3 is proceeded: Scenario 1: Price mechanism fails: Excessive resource set exists. Furthermore, the influence coefficients of all resources in this set are zero, i.e. Price updates cannot resolve resource conflicts; Scenario 2: Iteration Enters a Loop: Continuous In this iteration, the set of excess resources, the set of users using excess resources, and the temporary total reward remain consistent. The algorithm oscillates and cannot converge autonomously; Scenario 3: Reaching the computing power limit: The number of iterations reaches its maximum value. And there are still resources exceeding the limit; If the above scenario is not met, update the price of excess resources: ， Prices for resources not exceeding limits remain unchanged: ， Update iteration count Return to step 2.3 to execute the next iteration.

7. The low-Earth orbit satellite handover method with multi-user collaborative optimization as described in claim 1, characterized in that, Step 3 identifies users who occupy excessive resources as conflicting users, and the rest as non-conflicting users. The paths of non-conflicting users are fixed and their occupied resources are included in the disabled set. The non-cooperative game model of conflicting users is reconstructed. Price adjustment iteration is reproduced under the constraint of disabled resources. When the subgame converges to no resource excess, the optimal paths of the two types of users are merged to form a globally feasible solution.

8. The low-Earth orbit satellite handover method with multi-user collaborative optimization as described in claim 6, characterized in that, Step 3 specifically involves: Step 3.1: Divide users into conflicting users and non-conflicting users. The set of conflicting users consists of users who occupy at least one resource exceeding the limit, i.e.: ， in, Excessive resources when stopping iteration , To use excessive resources The user set, The union operation is performed on sets; the set of non-conflicting users is defined as the complement of the set of conflicting users. ， The path of a non-conflicting user at the point of stopping iteration is identified as the final path. This set of final paths for non-conflicting users is denoted as: And extract all satellite time slot resources occupied by these paths to form a set of disabled resources: ； Step 3.2: Reconstruct the non-cooperative game model of conflicting users: ， Gather the players into Remove all paths containing disabled resources to create new action space for conflicting users. The corresponding change to the profit function is as follows: Set the maximum number of iterations. Initialize the iteration counter Reconstructing the satellite time slot resource price matrix And set the initial price for conflicting user iterations: ， in, Under this setting, for Execute the iterative process in step 2.

3. When the iteration converges, that is... At that time, the set of optimal paths for conflicting users is: ， The fixed paths of non-conflicting users are merged with the convergence paths of conflicting users to form the final set of switching paths for all users: ， Through users Selected path Calculate the reward for the corresponding path And calculate the total reward for all user paths. : 。 9. A multi-user collaborative optimization low-Earth orbit satellite handover system, characterized in that, To implement the method as described in any one of claims 1 to 8, the following modules are included: Game model construction module: used to build non-cooperative game models, define the payoff functions of game players, action space and fusion path rewards and resource price costs, and form a multi-user switching decision optimization framework; Price Adjustment Iterative Optimization Module: Used to initialize satellite time slot resource prices and iterative parameters, solve the optimal user path using dynamic programming backtracking method, count resource occupancy and update over-limit resource prices, and perform iterative optimization; The conflict resolution and secondary optimization module is used to distinguish between conflicting and non-conflicting users, fix the paths and resources occupied by non-conflicting users, reconstruct the game model of conflicting users and complete the secondary iteration, and output the globally optimal switching solution.

Images